Research Publications
Browse peer-reviewed journal articles, conference papers, and scholarly publications. Use the filters below to refine results by type, author, year, or research area.
1.
Keerthi, H. K.; Rajini, S.; Karande, M. U.; Preethi, S.; Chaithra, M. H.; Sohi, Kaur
Reputation-Based Peer Selection in Decentralized Networks Using Distributed Federated Learning Models Proceedings
Institute of Electrical and Electronics Engineers Inc., 2025, ISBN: 9798331536770 (ISBN), (0).
@proceedings{728,
title = {Reputation-Based Peer Selection in Decentralized Networks Using Distributed Federated Learning Models},
author = {H. K. Keerthi and S. Rajini and M. U. Karande and S. Preethi and M. H. Chaithra and Kaur Sohi},
url = {https://ieeexplore.ieee.org/document/11210945},
doi = {10.1109/IACIS65746.2025.11210945},
isbn = {9798331536770 (ISBN)},
year = {2025},
date = {2025-01-01},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {Decentralized networks face significant challenges in peer selection due to the absence of centralized trust mechanisms, leading to vulnerabilities from malicious nodes and suboptimal resource utilization. Traditional federated learning approaches struggle with data heterogeneity and Byzantine attacks, compromising model integrity and convergence efficiency. This study proposes a novel reputationbased peer selection framework integrated with distributed federated learning models, employing dual-reputation computation schemes (debit-credit and credit-only) to evaluate peer contributions objectively. The system implements a layered architecture combining blockchain-based reputation storage with adaptive aggregation algorithms, enabling dynamic peer ranking and selective participation in model training rounds. Experimental validation demonstrates significant improvements in model accuracy (94.7% vs. 87.2% baseline), Byzantine fault tolerance (withstanding up to 35% malicious nodes), and communication efficiency (42% reduction in network overhead). The reputation system achieved 96.3% accuracy in malicious peer detection while maintaining 89.1% model convergence rate under heterogeneous data distributions. The proposed framework effectively addresses trust and security challenges in decentralized federated learning environments, providing robust peer selection mechanisms that enhance overall system performance and reliability while preserving data privacy and enabling scalable distributed machine learning applications.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
Decentralized networks face significant challenges in peer selection due to the absence of centralized trust mechanisms, leading to vulnerabilities from malicious nodes and suboptimal resource utilization. Traditional federated learning approaches struggle with data heterogeneity and Byzantine attacks, compromising model integrity and convergence efficiency. This study proposes a novel reputationbased peer selection framework integrated with distributed federated learning models, employing dual-reputation computation schemes (debit-credit and credit-only) to evaluate peer contributions objectively. The system implements a layered architecture combining blockchain-based reputation storage with adaptive aggregation algorithms, enabling dynamic peer ranking and selective participation in model training rounds. Experimental validation demonstrates significant improvements in model accuracy (94.7% vs. 87.2% baseline), Byzantine fault tolerance (withstanding up to 35% malicious nodes), and communication efficiency (42% reduction in network overhead). The reputation system achieved 96.3% accuracy in malicious peer detection while maintaining 89.1% model convergence rate under heterogeneous data distributions. The proposed framework effectively addresses trust and security challenges in decentralized federated learning environments, providing robust peer selection mechanisms that enhance overall system performance and reliability while preserving data privacy and enabling scalable distributed machine learning applications.
2.
Kumar, G. S.; Cheriyan, J.; Aparna, N.; Swathy, J.
Unleashing Facial Expression Recognition for Stress Detection Using Deep CNN Model. Proceedings
Elsevier B.V., vol. 259, 2025, ISBN: 18770509 (ISSN), (0).
@proceedings{441,
title = {Unleashing Facial Expression Recognition for Stress Detection Using Deep CNN Model.},
author = {G. S. Kumar and J. Cheriyan and N. Aparna and J. Swathy},
url = {https://www.sciencedirect.com/science/article/pii/S1877050925010762?via%3Dihub},
doi = {10.1016/j.procs.2025.03.332},
isbn = {18770509 (ISSN)},
year = {2025},
date = {2025-01-01},
journal = {Procedia Computer Science},
volume = {259},
pages = {306-315,},
publisher = {Elsevier B.V.},
abstract = {The technology and field of study known as Facial Emotion Recognition (FER) focuses on recognising and deciphering human facial expressions to ascertain emotions. It incorporates aspects of psychology, machine learning, and computer vision. Considering its ability to analyse facial expressions and spot signs of stress or emotional distress, FER plays a vital role in the detection of stress. Microexpressions are fleeting, uncontrollably expressed facial expressions that convey true feelings. These ephemeral expressions, which are frequently linked to stress, can be captured by FER systems. With its capacity to manage intricate patterns and variances in face expressions, FER utilising deep learning has grown in popularity. A novel FER system was designed to overcome issues that plague current FER systems, such as imbalanced datasets and robustness to noisy inputs. Large datasets are handled effectively by FER employing Deep CNN because of its architecture, which is built for high-dimensional input like images. Leveraging FER using Deep CNN achieves notable performance improvements, attaining an accuracy of 95.65 % and F1 score of 94.02%. Deep learning-based FER marks a substantial advancement in our ability to recognize and decipher facial expressions that convey emotions. FER technology has the ability to revolutionize several fields by improving intuition and sensitivity to human emotions through further development and thoughtful assessment of ethical ramifications. FER systems will advance in sophistication as science and technology develop, offering more profound understanding of human emotional states and promoting improved human-machine interactions.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
The technology and field of study known as Facial Emotion Recognition (FER) focuses on recognising and deciphering human facial expressions to ascertain emotions. It incorporates aspects of psychology, machine learning, and computer vision. Considering its ability to analyse facial expressions and spot signs of stress or emotional distress, FER plays a vital role in the detection of stress. Microexpressions are fleeting, uncontrollably expressed facial expressions that convey true feelings. These ephemeral expressions, which are frequently linked to stress, can be captured by FER systems. With its capacity to manage intricate patterns and variances in face expressions, FER utilising deep learning has grown in popularity. A novel FER system was designed to overcome issues that plague current FER systems, such as imbalanced datasets and robustness to noisy inputs. Large datasets are handled effectively by FER employing Deep CNN because of its architecture, which is built for high-dimensional input like images. Leveraging FER using Deep CNN achieves notable performance improvements, attaining an accuracy of 95.65 % and F1 score of 94.02%. Deep learning-based FER marks a substantial advancement in our ability to recognize and decipher facial expressions that convey emotions. FER technology has the ability to revolutionize several fields by improving intuition and sensitivity to human emotions through further development and thoughtful assessment of ethical ramifications. FER systems will advance in sophistication as science and technology develop, offering more profound understanding of human emotional states and promoting improved human-machine interactions.
3.
Jacob, Y. J.; Janney, B. J.; Hemalatha, RJ.; Preethi, S
Optimised hybrid deep learning classification model for kidney stone diagnosis Journal Article
In: Results in Engineering, vol. 26, 2025, ISBN: 25901230, (0).
@article{444,
title = {Optimised hybrid deep learning classification model for kidney stone diagnosis},
author = {Y. J. Jacob and B. J. Janney and RJ. Hemalatha and S Preethi},
url = {https://www.sciencedirect.com/science/article/pii/S2590123025012939?via%3Dihub},
doi = {10.1016/j.rineng.2025.105221},
isbn = {25901230},
year = {2025},
date = {2025-01-01},
journal = {Results in Engineering},
volume = {26},
publisher = {Elsevier B.V.},
abstract = {The kidney plays a vital role in maintaining homeostasis within the human body. In recent years, the prevalence of nephrolithiasis (kidney stone formation) characterized by the accumulation of crystalline solids within the renal system has emerged as a significant health concern. Early detection is critical for effective treatment and prevention of complications. Diagnostic imaging techniques such as computed tomography (CT), ultrasonography, and Doppler imaging are routinely employed for this purpose. To enhance the precision and reliability of early diagnosis, Deep Learning (DL) models are increasingly being integrated into the diagnostic workflow, offering superior accuracy through advanced image analysis and pattern recognition capabilities. The proposed work combines two deep learning models, AlexNet and Gated Recurrent Unit (GRU) for feature extraction and classification. These models are integrated to deliver optimal training parameter performance. An optimized AlexNet-GRU model is introduced in this work for detection of kidney stone, feature extraction, and classification. The Elephant Herding Optimizer (EHO) is utilized to fine-tune the hyperparameters of the AlexNet-GRU model. by performing this EHO fine tuning, the performance metrics of the proposed work have provided a high optimal result. Finally, the proposed evaluation metrics like precision, recall, accuracy, and F1 score are evaluated and compared with the traditional models to prove their efficient performances. The proposed model achieved a precision of 98.67 %, a recall of 97.68 %, an accuracy of 98.82 %, and an F1 score of 97.54 %.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
The kidney plays a vital role in maintaining homeostasis within the human body. In recent years, the prevalence of nephrolithiasis (kidney stone formation) characterized by the accumulation of crystalline solids within the renal system has emerged as a significant health concern. Early detection is critical for effective treatment and prevention of complications. Diagnostic imaging techniques such as computed tomography (CT), ultrasonography, and Doppler imaging are routinely employed for this purpose. To enhance the precision and reliability of early diagnosis, Deep Learning (DL) models are increasingly being integrated into the diagnostic workflow, offering superior accuracy through advanced image analysis and pattern recognition capabilities. The proposed work combines two deep learning models, AlexNet and Gated Recurrent Unit (GRU) for feature extraction and classification. These models are integrated to deliver optimal training parameter performance. An optimized AlexNet-GRU model is introduced in this work for detection of kidney stone, feature extraction, and classification. The Elephant Herding Optimizer (EHO) is utilized to fine-tune the hyperparameters of the AlexNet-GRU model. by performing this EHO fine tuning, the performance metrics of the proposed work have provided a high optimal result. Finally, the proposed evaluation metrics like precision, recall, accuracy, and F1 score are evaluated and compared with the traditional models to prove their efficient performances. The proposed model achieved a precision of 98.67 %, a recall of 97.68 %, an accuracy of 98.82 %, and an F1 score of 97.54 %.
4.
Raja, R.; Sureshkumar, K.; Rao, K. V.; Jayashree, N.
In: Energy Storage, vol. 7, 2025, ISBN: 25784862 (ISSN), (0).
@article{453,
title = {A Hybrid Approach for Smart Energy Management in Microgrids With Electric Vehicle Charging Using Snow Ablation Optimization and Cascade Chaotic Neural Network},
author = {R. Raja and K. Sureshkumar and K. V. Rao and N. Jayashree},
url = {https://onlinelibrary.wiley.com/doi/10.1002/est2.70208},
doi = {10.1002/est2.70208},
isbn = {25784862 (ISSN)},
year = {2025},
date = {2025-01-01},
journal = {Energy Storage},
volume = {7},
publisher = {John Wiley and Sons Inc},
abstract = {Integration of Renewable Energy Sources (RES) with Electric Vehicles (EVs) elucidates a crucial area in Energy Management (EM) for Microgrids (MGs). Probably the most difficult job is stochastic behavior from RES together with unpredictable EV charging demands, aspires towards grid stability, and destabilizes prompt frequency control. This article introduces a hybrid methodology designed for intelligent EM in MGs with EV charging. Proposed method integrates Snow Ablation Optimization (SAO) and Cascade Chaotic Neural Network (CCNN); therefore, it is called the SAO-CCNN technique. The aim is to improve economic performance of the MG integrated by EV charging by minimize the Operating Cost. SAO optimizes the utilization of RES and EVs, improving overall energy management. The CCNN is employed to predict the participation probability of EVs in grid support activities, thereby aiding in the accurate forecasting of energy demand. The suggested SAO-CCNN technique is implemented on MATLAB platform and evaluated against existing optimization methods, including Firefly Optimization Algorithm (FOA), Particle Swarm Optimization (PSO), Robust Optimization Algorithm (ROA), Multi Objective Optimization (MOO), and Whale Optimization Algorithm (WOA). The operating cost achieved using the proposed method is $17 184.1, demonstrating improved cost-efficiency compared to optimization methods.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
Integration of Renewable Energy Sources (RES) with Electric Vehicles (EVs) elucidates a crucial area in Energy Management (EM) for Microgrids (MGs). Probably the most difficult job is stochastic behavior from RES together with unpredictable EV charging demands, aspires towards grid stability, and destabilizes prompt frequency control. This article introduces a hybrid methodology designed for intelligent EM in MGs with EV charging. Proposed method integrates Snow Ablation Optimization (SAO) and Cascade Chaotic Neural Network (CCNN); therefore, it is called the SAO-CCNN technique. The aim is to improve economic performance of the MG integrated by EV charging by minimize the Operating Cost. SAO optimizes the utilization of RES and EVs, improving overall energy management. The CCNN is employed to predict the participation probability of EVs in grid support activities, thereby aiding in the accurate forecasting of energy demand. The suggested SAO-CCNN technique is implemented on MATLAB platform and evaluated against existing optimization methods, including Firefly Optimization Algorithm (FOA), Particle Swarm Optimization (PSO), Robust Optimization Algorithm (ROA), Multi Objective Optimization (MOO), and Whale Optimization Algorithm (WOA). The operating cost achieved using the proposed method is $17 184.1, demonstrating improved cost-efficiency compared to optimization methods.
5.
Lakshmi, D. V.; Madhusudhan, A.; Muniyandy, E.; Preethi, S.
In: Biomedical Materials and Devices, 2025, ISBN: 27314812 (ISSN), (0).
@article{455,
title = {Return-Aligned Random Graph Diffusion with Dual-Channel Temporal Convolutional Network-Based Classification of Epithelial Ovarian Cancer on T2W-MRI},
author = {D. V. Lakshmi and A. Madhusudhan and E. Muniyandy and S. Preethi},
url = {https://link.springer.com/article/10.1007/s44174-025-00381-7},
doi = {10.1007/s44174-025-00381-7},
isbn = {27314812 (ISSN)},
year = {2025},
date = {2025-01-01},
journal = {Biomedical Materials and Devices},
publisher = {Springer Nature},
abstract = {This study aims to develop a highly accurate and efficient deep-learning framework for the automated classification of epithelial ovarian cancer (EOC) subtypes using T2-weighted MRI (T2W-MRI) images. The objective is to overcome limitations such as poor contrast, high inter-class variation, dataset imbalance, and computational complexity that hinder current diagnostic methods. To address these, we propose the return-aligned random graph diffusion with dual-channel temporal convolutional network (RA-RGD-DCTCNet) model, evaluated on the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) datasets. Image quality is first enhanced using Discrete Wavelet Transformation with Pre-Gaussian Filtering (DWT-PGF), followed by precise tumor segmentation via the return-aligned decision transformer (RADT). The random graph diffusion dual-channel temporal convolutional network (RGD-DCTCNet) performs feature extraction and classification, with accuracy further boosted by the Secretary Bird Optimization Algorithm (SBOA). Experimental results demonstrate that the RA-RGD-DCTCNet model achieves 99.9% accuracy and 99.8% sensitivity, significantly outperforming existing methods and showing promise for clinical application in reliable, automated cancer diagnosis.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
This study aims to develop a highly accurate and efficient deep-learning framework for the automated classification of epithelial ovarian cancer (EOC) subtypes using T2-weighted MRI (T2W-MRI) images. The objective is to overcome limitations such as poor contrast, high inter-class variation, dataset imbalance, and computational complexity that hinder current diagnostic methods. To address these, we propose the return-aligned random graph diffusion with dual-channel temporal convolutional network (RA-RGD-DCTCNet) model, evaluated on the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) datasets. Image quality is first enhanced using Discrete Wavelet Transformation with Pre-Gaussian Filtering (DWT-PGF), followed by precise tumor segmentation via the return-aligned decision transformer (RADT). The random graph diffusion dual-channel temporal convolutional network (RGD-DCTCNet) performs feature extraction and classification, with accuracy further boosted by the Secretary Bird Optimization Algorithm (SBOA). Experimental results demonstrate that the RA-RGD-DCTCNet model achieves 99.9% accuracy and 99.8% sensitivity, significantly outperforming existing methods and showing promise for clinical application in reliable, automated cancer diagnosis.
6.
Barik, B.; Sairam, M. V. S.; Naresh, N.; Preethi, S.
In: Transactions on Electrical and Electronic Materials, 2025, ISBN: 12297607 (ISSN), (0).
@article{471,
title = {Channel Estimation for Massive MIMO-OFDM Systems Using Heterogeneous Edge-Enhanced Graph Hamiltonian Quantum Generative Adversarial Networks with Imperfect Channel State Information},
author = {B. Barik and M. V. S. Sairam and N. Naresh and S. Preethi},
url = {https://link.springer.com/article/10.1007/s42341-025-00649-1},
doi = {10.1007/s42341-025-00649-1},
isbn = {12297607 (ISSN)},
year = {2025},
date = {2025-01-01},
journal = {Transactions on Electrical and Electronic Materials},
publisher = {Korean Institute of Electrical and Electronic Material Engineers},
abstract = {Massive MIMO-OFDM systems are integral to next-generation wireless communication networks due to their ability to achieve high spectral efficiency and data rates. However, accurate channel estimation in these systems is challenging, particularly when channel state information (CSI) is imperfect. Traditional methods often fail to generalize across varying environments or struggle with high computational complexity, leading to suboptimal performance. This paper introduces a novel framework Heterogeneous Edge-Enhanced Graph Hamiltonian Quantum Generative Adversarial Networks (Hedge-GAN) for channel estimation in Massive MIMO-OFDM systems. The Hedge-GAN component effectively models the probabilistic nature of imperfect CSI, leveraging quantum computation for enhanced expressivity and efficiency. The Hedge-GAN component captures the spatial and temporal dependencies in the system, ensuring robust feature extraction even under heterogeneous network conditions. Existing solutions suffer from limited scalability, inadequate utilization of spatial information, and subpar optimization of network parameters. To address these, the proposed framework integrates the Musical Chairs Optimization Algorithm (MCOA), a bio-inspired metaheuristic, to optimize Hedge-GAN’s hyperparameters. This integration enhances convergence speed and improves estimation accuracy. Comparative analysis against state-of-the-art methods demonstrates that the proposed approach significantly improves performance in terms of mean squared error (0.40), spectral efficiency (8.6bps/Hz), and robustness to imperfect CSI, paving the way for more reliable and efficient wireless communication systems.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
Massive MIMO-OFDM systems are integral to next-generation wireless communication networks due to their ability to achieve high spectral efficiency and data rates. However, accurate channel estimation in these systems is challenging, particularly when channel state information (CSI) is imperfect. Traditional methods often fail to generalize across varying environments or struggle with high computational complexity, leading to suboptimal performance. This paper introduces a novel framework Heterogeneous Edge-Enhanced Graph Hamiltonian Quantum Generative Adversarial Networks (Hedge-GAN) for channel estimation in Massive MIMO-OFDM systems. The Hedge-GAN component effectively models the probabilistic nature of imperfect CSI, leveraging quantum computation for enhanced expressivity and efficiency. The Hedge-GAN component captures the spatial and temporal dependencies in the system, ensuring robust feature extraction even under heterogeneous network conditions. Existing solutions suffer from limited scalability, inadequate utilization of spatial information, and subpar optimization of network parameters. To address these, the proposed framework integrates the Musical Chairs Optimization Algorithm (MCOA), a bio-inspired metaheuristic, to optimize Hedge-GAN’s hyperparameters. This integration enhances convergence speed and improves estimation accuracy. Comparative analysis against state-of-the-art methods demonstrates that the proposed approach significantly improves performance in terms of mean squared error (0.40), spectral efficiency (8.6bps/Hz), and robustness to imperfect CSI, paving the way for more reliable and efficient wireless communication systems.
7.
Karthik, P.; Chethan, B.; Pavan, B.; Pateel, R.
A Novel Approach for an Enhanced Security System on IoT Devices Proceedings
2025.
@proceedings{613,
title = {A Novel Approach for an Enhanced Security System on IoT Devices},
author = {P. Karthik and B. Chethan and B. Pavan and R. Pateel},
url = {https://ieeexplore.ieee.org/document/11076637},
doi = {10.1109/ICMOCE64100.2025.11076637},
year = {2025},
date = {2025-01-01},
journal = {2025 International Conference on Microwave, Optical, and Communication Engineering (ICMOCE)},
pages = {1-6,},
abstract = {IoT devices increase the variety and complexity of security. Current security solutions typically fall short in a number of areas, especially when it comes to detecting and responding to attacks in real time. Through the interaction. Some examples are e-commerce, smart industrial management, smart grid, smart health, smart home, smart workplace, and e-governance. Every day, more and more devices get connected, which raises security risks and issues A comprehensive IoT security model is needed to support resource-based IoT gadgets and end security. We focus on the networks and IoT system applications in addition to the organizational approach towards IoT security, attack vectors, and security needs for IoT systems. We also provide a baseline for security implementation and recommend a security architecture to offer security-enabled IoT services through the the application of this cybersecurity framework, the project aims to strengthen the security stance of IoT devices, reduce the attack surface, and foster user trust in IoT applications. By establishing best practices and guidelines, this framework serves as a foundational step toward creating a safer IoT environment, ultimately contributing to the broader goals of digital Privacy and security in and an interconnected world. Experimental results demonstrate the effectiveness of AI/ML techniques in decreasing down on false positives and enhancing response times, making IoT deployments more resilient to emerging security threats.},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
IoT devices increase the variety and complexity of security. Current security solutions typically fall short in a number of areas, especially when it comes to detecting and responding to attacks in real time. Through the interaction. Some examples are e-commerce, smart industrial management, smart grid, smart health, smart home, smart workplace, and e-governance. Every day, more and more devices get connected, which raises security risks and issues A comprehensive IoT security model is needed to support resource-based IoT gadgets and end security. We focus on the networks and IoT system applications in addition to the organizational approach towards IoT security, attack vectors, and security needs for IoT systems. We also provide a baseline for security implementation and recommend a security architecture to offer security-enabled IoT services through the the application of this cybersecurity framework, the project aims to strengthen the security stance of IoT devices, reduce the attack surface, and foster user trust in IoT applications. By establishing best practices and guidelines, this framework serves as a foundational step toward creating a safer IoT environment, ultimately contributing to the broader goals of digital Privacy and security in and an interconnected world. Experimental results demonstrate the effectiveness of AI/ML techniques in decreasing down on false positives and enhancing response times, making IoT deployments more resilient to emerging security threats.
8.
Singh, S.; Omar, M.; Sumathi, D.; P, K.
Federated Learning-Based Multi-Objective Optimization for IoT-Enabled Distributed Environmental Monitoring in Consumer Electronics Journal Article
In: IEEE Transactions on Consumer Electronics, pp. 1-1,, 2025, ISBN: 1558-4127.
@article{614,
title = {Federated Learning-Based Multi-Objective Optimization for IoT-Enabled Distributed Environmental Monitoring in Consumer Electronics},
author = {S. Singh and M. Omar and D. Sumathi and K. P},
url = {https://ieeexplore.ieee.org/document/11105555},
doi = {10.1109/TCE.2025.3594404},
isbn = {1558-4127},
year = {2025},
date = {2025-01-01},
journal = {IEEE Transactions on Consumer Electronics},
pages = {1-1,},
abstract = {In this paper, the integration of the Internet of Things (IoT) in consumer electronics has significantly improved the ability to monitor environmental conditions in real time. This advancement is particularly crucial in patient care and medical data management. However, the vast volume of data generated by these devices demands advanced optimization algorithms to efficiently manage and evaluate this information. This research proposes a next-generation environmental monitoring system designed with a novel Federated Learning-Based Multi-Objective Optimization (FL-MOO) algorithm. This algorithm introduces specific learning parameters designed to enhance data processing quality, ensuring real-time monitoring with minimal latency and optimal resource allocation. By balancing the computational load and maintaining the accuracy of the data across devices enabled by distributed IoT, the proposed FL-MOO algorithm achieves superior performance in terms of speed, accuracy, and resource utilization compared to conventional methods. Simulation experiments demonstrate the efficiency of the proposed algorithm in optimizing data processing within a distributed environment, effectively managing the load distribution between devices. The algorithm also significantly reduced environmental overhead, achieving an average energy consumption of 0.35145, a latency of 0.17376, and a load balancing value of 0.43452. This work paves the way for more scalable and powerful IoT-based solutions in smart health, driving advances in healthcare delivery, patient outcomes, and the effective integration of consumer electronics in environmental monitoring systems.},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
In this paper, the integration of the Internet of Things (IoT) in consumer electronics has significantly improved the ability to monitor environmental conditions in real time. This advancement is particularly crucial in patient care and medical data management. However, the vast volume of data generated by these devices demands advanced optimization algorithms to efficiently manage and evaluate this information. This research proposes a next-generation environmental monitoring system designed with a novel Federated Learning-Based Multi-Objective Optimization (FL-MOO) algorithm. This algorithm introduces specific learning parameters designed to enhance data processing quality, ensuring real-time monitoring with minimal latency and optimal resource allocation. By balancing the computational load and maintaining the accuracy of the data across devices enabled by distributed IoT, the proposed FL-MOO algorithm achieves superior performance in terms of speed, accuracy, and resource utilization compared to conventional methods. Simulation experiments demonstrate the efficiency of the proposed algorithm in optimizing data processing within a distributed environment, effectively managing the load distribution between devices. The algorithm also significantly reduced environmental overhead, achieving an average energy consumption of 0.35145, a latency of 0.17376, and a load balancing value of 0.43452. This work paves the way for more scalable and powerful IoT-based solutions in smart health, driving advances in healthcare delivery, patient outcomes, and the effective integration of consumer electronics in environmental monitoring systems.
9.
Sivaramkrishnan, M.; R, B.; Kavitha, D.; Emayavaramban, G.; Abu-Saleem, Ahmad; Chadge, R.
2025.
@proceedings{615,
title = {Optimized Energy-Efficient Speed Control in Lightweight Electric Vehicle using Dynamic Spiking Graph Neural Network},
author = {M. Sivaramkrishnan and B. R and D. Kavitha and G. Emayavaramban and Ahmad Abu-Saleem and R. Chadge},
url = {https://ieeexplore.ieee.org/document/11073910},
doi = {10.1109/ICIMA64861.2025.11073910},
year = {2025},
date = {2025-01-01},
journal = {2025 7th International Conference on Inventive Material Science and Applications (ICIMA)},
pages = {617-622,},
abstract = {Efficient Electric Vehicle (EV) speed control ensures smooth acceleration, stability, and energy optimization under different driving conditions. The speed control system in lightweight EVs enables effective velocity regulation which improves both performance characteristics and energy optimization during driving operations. In lightweight EVs speed control operates with high sensitivity to external disturbances such as road inclines together with load variations but this result in inefficient energy utilization. In order to address these issues, this paper proposes an approach of Dynamic Spiking Graph Neural Network (DSGNN) for speed control in light weight EV. The main goal is to improve energy efficiency in lightweight EVs by optimizing speed control using DSGNN. DSGNN is utilized to predict speed variations in lightweight EVs, capturing temporal dependencies for accurate speed tracking and control. The proposed method undergoes implementation and evaluation using MATLAB against various existing approaches, such as Ultra-Local Model-Based Fuzzy QLearning Multi-Agent System (ULM-FQMAS), Ant Colony Optimization- Fractional-Order Proportional Integral Derivative (ACO-FOPID), Fuzzy Proportional Integral Derivative (FPID), Self-Constructing Type-2 Fuzzy Neural Network (SCT2FNN) and Adaptive Neuro-Fuzzy Model Predictive Control (ANFMPC). The proposed DSGNN method delivers improved energy efficiency by 29.21% which proves its effectiveness for optimizing speed control and power utilization in lightweight EVs.},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
Efficient Electric Vehicle (EV) speed control ensures smooth acceleration, stability, and energy optimization under different driving conditions. The speed control system in lightweight EVs enables effective velocity regulation which improves both performance characteristics and energy optimization during driving operations. In lightweight EVs speed control operates with high sensitivity to external disturbances such as road inclines together with load variations but this result in inefficient energy utilization. In order to address these issues, this paper proposes an approach of Dynamic Spiking Graph Neural Network (DSGNN) for speed control in light weight EV. The main goal is to improve energy efficiency in lightweight EVs by optimizing speed control using DSGNN. DSGNN is utilized to predict speed variations in lightweight EVs, capturing temporal dependencies for accurate speed tracking and control. The proposed method undergoes implementation and evaluation using MATLAB against various existing approaches, such as Ultra-Local Model-Based Fuzzy QLearning Multi-Agent System (ULM-FQMAS), Ant Colony Optimization- Fractional-Order Proportional Integral Derivative (ACO-FOPID), Fuzzy Proportional Integral Derivative (FPID), Self-Constructing Type-2 Fuzzy Neural Network (SCT2FNN) and Adaptive Neuro-Fuzzy Model Predictive Control (ANFMPC). The proposed DSGNN method delivers improved energy efficiency by 29.21% which proves its effectiveness for optimizing speed control and power utilization in lightweight EVs.
10.
Balakrishna, K. K.; Karthik, K.
High-Speed Fiber-Optic Communication Performance Utilizing Fiber Bragg Grating-Based Dispersion Compensation Schemes Journal Article
In: International Journal of Advanced Computer Science and Applications, vol. 16, pp. 210-220,, 2025, ISBN: 21565570 (ISSN); 2158107X (ISSN), (0).
@article{629,
title = {High-Speed Fiber-Optic Communication Performance Utilizing Fiber Bragg Grating-Based Dispersion Compensation Schemes},
author = {K. K. Balakrishna and K. Karthik},
url = {https://thesai.org/Publications/ViewPaper?Volume=16&Issue=7&Code=ijacsa&SerialNo=22},
doi = {10.14569/IJACSA.2025.0160722},
isbn = {21565570 (ISSN); 2158107X (ISSN)},
year = {2025},
date = {2025-01-01},
journal = {International Journal of Advanced Computer Science and Applications},
volume = {16},
pages = {210-220,},
publisher = {Science and Information Organization},
abstract = {Chromatic dispersion is a significant limitation in optical fiber communication, as it causes pulse broadening, which negatively impacts transmission distance and data rates, both of which are critical for meeting the high-speed demands of 5G optical networks. This study focuses on addressing chromatic dispersion in Standard Single-Mode Fiber (SSMF) systems, which are widely deployed in 5G fronthaul and access networks. A comprehensive investigation is conducted using Gaussian-apodized linear chirped Fiber Bragg Gratings (FBGs) for dispersion compensation, implemented across three strategic configurations: pre-compensation, post-compensation, and symmetrical compensation. Each scheme is systematically evaluated to determine the most effective approach for enhancing signal integrity and overall network performance. Simulations are performed using OptiSystem 7.0 on a 10 Gbps SSMF-based optical system, with transmission distances ranging from 10 km to 80 km under controlled simulation parameters. Key performance metrics, including Quality factor (Q-factor), Bit Error Rate (BER), and eye height, are analyzed by varying SSMF length, input power, and bit rate. The results demonstrate that symmetrical compensation using Gaussian-apodized linear chirped FBGs provides the best performance, achieving a Q-factor of 12.3938, an ultra-low BER of 1.12336×10-35, and a significantly improved eye height at 80 km. These findings establish the symmetrical compensation scheme employing Apodized Chirped Fiber Bragg Gratings (ACFBGs) as the most effective and scalable solution for high-speed, long-distance optical transmission in 5G networks. This approach enables key 5G applications, including ultra-reliable low-latency communication (URLLC), enhanced mobile broadband (eMBB), and smart infrastructure in smart cities. The proposed technique offers multiple advantages, such as low BER, high Q-factor, reduced signal distortion through sidelobe suppression, energy efficiency via passive operation, and design flexibility for long-haul network integration.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
Chromatic dispersion is a significant limitation in optical fiber communication, as it causes pulse broadening, which negatively impacts transmission distance and data rates, both of which are critical for meeting the high-speed demands of 5G optical networks. This study focuses on addressing chromatic dispersion in Standard Single-Mode Fiber (SSMF) systems, which are widely deployed in 5G fronthaul and access networks. A comprehensive investigation is conducted using Gaussian-apodized linear chirped Fiber Bragg Gratings (FBGs) for dispersion compensation, implemented across three strategic configurations: pre-compensation, post-compensation, and symmetrical compensation. Each scheme is systematically evaluated to determine the most effective approach for enhancing signal integrity and overall network performance. Simulations are performed using OptiSystem 7.0 on a 10 Gbps SSMF-based optical system, with transmission distances ranging from 10 km to 80 km under controlled simulation parameters. Key performance metrics, including Quality factor (Q-factor), Bit Error Rate (BER), and eye height, are analyzed by varying SSMF length, input power, and bit rate. The results demonstrate that symmetrical compensation using Gaussian-apodized linear chirped FBGs provides the best performance, achieving a Q-factor of 12.3938, an ultra-low BER of 1.12336×10-35, and a significantly improved eye height at 80 km. These findings establish the symmetrical compensation scheme employing Apodized Chirped Fiber Bragg Gratings (ACFBGs) as the most effective and scalable solution for high-speed, long-distance optical transmission in 5G networks. This approach enables key 5G applications, including ultra-reliable low-latency communication (URLLC), enhanced mobile broadband (eMBB), and smart infrastructure in smart cities. The proposed technique offers multiple advantages, such as low BER, high Q-factor, reduced signal distortion through sidelobe suppression, energy efficiency via passive operation, and design flexibility for long-haul network integration.
11.
Karthik, P.; Rajesh, R.; Kumar, Nithin; Kumar, Kalyan; Kiran, S.
Content-Based Fashion Image Retrieval in Android Applications using Artificial Neural Networks Proceedings
Institute of Electrical and Electronics Engineers Inc., 2025, ISBN: 9798331512118 (ISBN), (0).
@proceedings{635,
title = {Content-Based Fashion Image Retrieval in Android Applications using Artificial Neural Networks},
author = {P. Karthik and R. Rajesh and Nithin Kumar and Kalyan Kumar and S. Kiran},
url = {https://ieeexplore.ieee.org/document/11140848},
doi = {10.1109/ICCMC65190.2025.11140848},
isbn = {9798331512118 (ISBN)},
year = {2025},
date = {2025-01-01},
pages = {1568-1573,},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {The rapid rise of e-commerce and digital fashion has increased demand for intelligent search systems. Traditional keyword-based approaches fail to accurately capture user preferences, often leading to poor recommendations. This research addresses this challenge by proposing a Content-Based Image Retrieval (CBIR) system tailored for fashion, deployed in a React-Native mobile application. The system uses deep learning models, including Convolutional Neural Networks (CNNs) and Artificial Neural Networks (ANNs), to extract detailed features from fashion images. These features are indexed and compared using similarity metrics like cosine similarity and Euclidean distance to retrieve relevant results. Autoencoders and attention mechanisms refine feature extraction, while contrastive and triplet loss functions improve embedding quality. Users can search by uploading fashion images, leading to a more intuitive shopping experience. The solution demonstrates high effectiveness, achieving a stable training and validation accuracy of approximately 95.99% across multiple epochs. This study proves the viability of deep learning-powered CBIR for fashion, improving mobile-based recommendations and offering reliable real-time results. The integration of efficient retrieval techniques, optimized indexing, and ANN-based models ensures the system is scalable and user-friendly for practical applications in fashion search.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
The rapid rise of e-commerce and digital fashion has increased demand for intelligent search systems. Traditional keyword-based approaches fail to accurately capture user preferences, often leading to poor recommendations. This research addresses this challenge by proposing a Content-Based Image Retrieval (CBIR) system tailored for fashion, deployed in a React-Native mobile application. The system uses deep learning models, including Convolutional Neural Networks (CNNs) and Artificial Neural Networks (ANNs), to extract detailed features from fashion images. These features are indexed and compared using similarity metrics like cosine similarity and Euclidean distance to retrieve relevant results. Autoencoders and attention mechanisms refine feature extraction, while contrastive and triplet loss functions improve embedding quality. Users can search by uploading fashion images, leading to a more intuitive shopping experience. The solution demonstrates high effectiveness, achieving a stable training and validation accuracy of approximately 95.99% across multiple epochs. This study proves the viability of deep learning-powered CBIR for fashion, improving mobile-based recommendations and offering reliable real-time results. The integration of efficient retrieval techniques, optimized indexing, and ANN-based models ensures the system is scalable and user-friendly for practical applications in fashion search.
12.
Sapna, R.; Vijayalaxmi, Y.; Kavya, V. R.; Manasa, L.
Efficient Restaurant Management System Through Machine Learning and NLP-Based Sentiment Analysis Proceedings
Institute of Electrical and Electronics Engineers Inc., 2025, ISBN: 9798331531034 (ISBN), (0).
@proceedings{642,
title = {Efficient Restaurant Management System Through Machine Learning and NLP-Based Sentiment Analysis},
author = {R. Sapna and Y. Vijayalaxmi and V. R. Kavya and L. Manasa},
url = {https://ieeexplore.ieee.org/document/11140044},
doi = {10.1109/INCET64471.2025.11140044},
isbn = {9798331531034 (ISBN)},
year = {2025},
date = {2025-01-01},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {The rapid growth of online restaurant reviews has made sentiment analysis a crucial tool for understanding customer feedback and enhancing service quality. Restaurant quality including food and drinks, environment, place, and service directly impacts brand image and determines customer satisfaction. Traditional review evaluation methods are time-consuming and inefficient, creating a need for automated sentiment classification. This study shows an optimized restaurant management system that introduces machine learning (ML) and natural language processing (NLP) techniques to study customer sentiments. By evaluating these sentiments, businesses can obtain valuable observations into customer satisfaction and recognize areas of improvement. The methodology consists of data preprocessing, feature extraction, and selection of the best ML model from a set containing Decision Trees, Support Vector Machines, Random Forest, and XGBoost, to rank reviews into positive, negative, or neutral sentiments. Additionally, voice search and bulk testing functionalities enhance system usability. Experimental results display high classification accuracy, enabling restaurant owners to glean insightful information from customer reviews. By streamlining sentiment analysis, this system aids in data-driven decision-making, ultimately improving customer satisfaction and operational efficiency.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
The rapid growth of online restaurant reviews has made sentiment analysis a crucial tool for understanding customer feedback and enhancing service quality. Restaurant quality including food and drinks, environment, place, and service directly impacts brand image and determines customer satisfaction. Traditional review evaluation methods are time-consuming and inefficient, creating a need for automated sentiment classification. This study shows an optimized restaurant management system that introduces machine learning (ML) and natural language processing (NLP) techniques to study customer sentiments. By evaluating these sentiments, businesses can obtain valuable observations into customer satisfaction and recognize areas of improvement. The methodology consists of data preprocessing, feature extraction, and selection of the best ML model from a set containing Decision Trees, Support Vector Machines, Random Forest, and XGBoost, to rank reviews into positive, negative, or neutral sentiments. Additionally, voice search and bulk testing functionalities enhance system usability. Experimental results display high classification accuracy, enabling restaurant owners to glean insightful information from customer reviews. By streamlining sentiment analysis, this system aids in data-driven decision-making, ultimately improving customer satisfaction and operational efficiency.
13.
Raj, J. J.; Sasikala, T.; Ezhilarasan, K.; Rajesh, K.; Karthik, K.; Immanuel, J. S.
Adaptive Attribute-based Encryption(A-ABE) Framework for Securing Smart IoT Networks Proceedings
Institute of Electrical and Electronics Engineers Inc., 2025, ISBN: 9798331594916 (ISBN), (0).
@proceedings{683,
title = {Adaptive Attribute-based Encryption(A-ABE) Framework for Securing Smart IoT Networks},
author = {J. J. Raj and T. Sasikala and K. Ezhilarasan and K. Rajesh and K. Karthik and J. S. Immanuel},
url = {https://ieeexplore.ieee.org/document/11171299},
doi = {10.1109/ICSCSA66339.2025.11171299},
isbn = {9798331594916 (ISBN)},
year = {2025},
date = {2025-01-01},
pages = {249-254,},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {In this paper, we proposed and implemented an Adaptive Attribute-Based Encryption (A-ABE) Framework that combines Ciphertext-Policy ABE with context-aware machine learning to secure smart IoT networks. The framework focused on key limitations of traditional ABE schemes, and their inability for adapting to dynamic user contexts and access requirements in real time scenarios. By deploying lightweight machine learning models, the system can be enabled to intelligently analyze contextual data like location, user behavior, and role changes - and update access policies accordingly, without requiring manual re-encryption or administrative intervention. The implementation and evaluation conducted in a simulated smart healthcare environment demonstrated that the A-ABE framework offers an optimal balance between security, adaptability, and system efficiency. The experimental results showed low encryption and decryption delays, high policy enforcement accuracy , and modest increases in resource utilization, making the solution is viable for deployment on edge devices. Overall, the proposed framework improves both the confidentiality of sensitive IoT data and the resilience of access control mechanisms against insider threats, contextual anomalies, and unauthorized access. This research confirms that combining ABE with real-time machine learning provides a scalable and intelligent approach to enforcing secure access in dynamic IoT environments.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
In this paper, we proposed and implemented an Adaptive Attribute-Based Encryption (A-ABE) Framework that combines Ciphertext-Policy ABE with context-aware machine learning to secure smart IoT networks. The framework focused on key limitations of traditional ABE schemes, and their inability for adapting to dynamic user contexts and access requirements in real time scenarios. By deploying lightweight machine learning models, the system can be enabled to intelligently analyze contextual data like location, user behavior, and role changes - and update access policies accordingly, without requiring manual re-encryption or administrative intervention. The implementation and evaluation conducted in a simulated smart healthcare environment demonstrated that the A-ABE framework offers an optimal balance between security, adaptability, and system efficiency. The experimental results showed low encryption and decryption delays, high policy enforcement accuracy , and modest increases in resource utilization, making the solution is viable for deployment on edge devices. Overall, the proposed framework improves both the confidentiality of sensitive IoT data and the resilience of access control mechanisms against insider threats, contextual anomalies, and unauthorized access. This research confirms that combining ABE with real-time machine learning provides a scalable and intelligent approach to enforcing secure access in dynamic IoT environments.
14.
Raj, J. J.; Sasikala, T.; Ezhilarasan, K.; Rajesh, K.; Karthik, K.; Immanuel, J. S.
A Lightweight and Explainable Machine Learning Approach for Intrusion Detection Proceedings
Institute of Electrical and Electronics Engineers Inc., 2025, ISBN: 9798331594916 (ISBN), (0).
@proceedings{684,
title = {A Lightweight and Explainable Machine Learning Approach for Intrusion Detection},
author = {J. J. Raj and T. Sasikala and K. Ezhilarasan and K. Rajesh and K. Karthik and J. S. Immanuel},
url = {https://ieeexplore.ieee.org/document/11171135},
doi = {10.1109/ICSCSA66339.2025.11171135},
isbn = {9798331594916 (ISBN)},
year = {2025},
date = {2025-01-01},
pages = {243-248,},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {In the digital era, the detection of malicious network activities and an unauthorized and still remains a critical challenge. Intrusion Detection Systems (IDS) is a vital player in safeguarding digital infrastructures; Due to the advancement in technologies cyber-attacks necessitates the adoption of intelligent, data-driven approaches. This work explores the application of logistic regression, a statistical machine learning technique, for intrusion detection using the NSL-KDD dataset. The research emphasizes the model's efficiency, interpretability, and capability to differentiate normal and malicious traffic in binary classification tasks. Experimental results shows that the logistic regression achieves a detection accuracy of 91.6%, with strong precision and recall metrics, making it suitable for resource-constrained or real-time environments. Comparative analysis with alternative classifiers highlights logistic regression's computational advantages and transparency. The research reveals that, despite its simplicity, logistic regression presents a robust and explainable solution to mitigate intrusion detections and offers a foundation for further enhancements through hybrid and adaptive modeling techniques .},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
In the digital era, the detection of malicious network activities and an unauthorized and still remains a critical challenge. Intrusion Detection Systems (IDS) is a vital player in safeguarding digital infrastructures; Due to the advancement in technologies cyber-attacks necessitates the adoption of intelligent, data-driven approaches. This work explores the application of logistic regression, a statistical machine learning technique, for intrusion detection using the NSL-KDD dataset. The research emphasizes the model's efficiency, interpretability, and capability to differentiate normal and malicious traffic in binary classification tasks. Experimental results shows that the logistic regression achieves a detection accuracy of 91.6%, with strong precision and recall metrics, making it suitable for resource-constrained or real-time environments. Comparative analysis with alternative classifiers highlights logistic regression's computational advantages and transparency. The research reveals that, despite its simplicity, logistic regression presents a robust and explainable solution to mitigate intrusion detections and offers a foundation for further enhancements through hybrid and adaptive modeling techniques .
15.
Sharma, M.; Chaturvedi, P.; Singh, N.; Tiwari, K.; Mohan, M.
Mushroom Classification using Transfer Learning Techniques Proceedings
Institute of Electrical and Electronics Engineers Inc., 2025, ISBN: 9798331544119 (ISBN), (0).
@proceedings{699,
title = {Mushroom Classification using Transfer Learning Techniques},
author = {M. Sharma and P. Chaturvedi and N. Singh and K. Tiwari and M. Mohan},
url = {https://ieeexplore.ieee.org/document/11189593},
doi = {10.1109/CIACON65473.2025.11189593},
isbn = {9798331544119 (ISBN)},
year = {2025},
date = {2025-01-01},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {Accidental intake of toxic mushrooms is still a very serious public health problem even in this modern age when mushrooms are consumed everywhere like food and medicine. Mycological identification has traditionally been dependent on expert knowledge derived from a visual appraisal of basic cap, gills, stem, and spore print characteristics; hence this is one of the reasons why, in many instances, edibles cannot be distinguished from poisonous mushrooms. These techniques can be error-prone and are often subjective. This study employs convolutional neural networks to facilitate the classification of mushrooms associated with image data for solving the problem. This study considers five classes that are often encountered from the dataset: "Mushrooms Classification - Common Genus's Images"from Kaggle. Transfer learning was based on ResNet50V2 architecture. Weights were pre-trained to enhance feature extraction and classification. Techniques of data augmentation were employed to enhance variability and robustness of the dataset. The model now possesses an accuracy of 81.41%, thereby proving its efficacy in discrimination among the species of mushrooms based on their observable features. This strategy has great potential to be successful with the creation of a scalable, economically feasible tool for safer mushroom foraging and consumption.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
Accidental intake of toxic mushrooms is still a very serious public health problem even in this modern age when mushrooms are consumed everywhere like food and medicine. Mycological identification has traditionally been dependent on expert knowledge derived from a visual appraisal of basic cap, gills, stem, and spore print characteristics; hence this is one of the reasons why, in many instances, edibles cannot be distinguished from poisonous mushrooms. These techniques can be error-prone and are often subjective. This study employs convolutional neural networks to facilitate the classification of mushrooms associated with image data for solving the problem. This study considers five classes that are often encountered from the dataset: "Mushrooms Classification - Common Genus's Images"from Kaggle. Transfer learning was based on ResNet50V2 architecture. Weights were pre-trained to enhance feature extraction and classification. Techniques of data augmentation were employed to enhance variability and robustness of the dataset. The model now possesses an accuracy of 81.41%, thereby proving its efficacy in discrimination among the species of mushrooms based on their observable features. This strategy has great potential to be successful with the creation of a scalable, economically feasible tool for safer mushroom foraging and consumption.
16.
Deepthi, C. V.; Bingi, L.; Shoma, R. S.; Arpitha, K.; Lohith, C.; Vasumathi, A. K.; Gnanasundari, A.
FreqPatchNet: A Dual-Domain Patch-Wise Fusion Network for Robust Phase Correction in Underwater Image Reconstruction Journal Article
In: Engineering, Technology and Applied Science Research, vol. 15, pp. 26771-26776,, 2025, ISBN: 22414487 (ISSN), (0).
@article{720,
title = {FreqPatchNet: A Dual-Domain Patch-Wise Fusion Network for Robust Phase Correction in Underwater Image Reconstruction},
author = {C. V. Deepthi and L. Bingi and R. S. Shoma and K. Arpitha and C. Lohith and A. K. Vasumathi and A. Gnanasundari},
url = {https://etasr.com/index.php/ETASR/article/view/12990},
doi = {10.48084/etasr.12990},
isbn = {22414487 (ISSN)},
year = {2025},
date = {2025-01-01},
journal = {Engineering, Technology and Applied Science Research},
volume = {15},
pages = {26771-26776,},
publisher = {Dr D. Pylarinos},
abstract = {This paper presents FreqPatchNet, a novel patch-wise dual-domain Convolutional Neural Network (CNN) designed to correct phase distortions in underwater images. The model uses bispectral frequency features and local CNN regression to reconstruct clean images from distorted inputs. Evaluated using Peak Signal-to-Noise Ratio (PSNR) and Mean Squared Error (MSE), FreqPatchNet achieves a maximum PSNR of 35.6 dB and a lowest MSE of 0.28 at 10% distortion. A comparative analysis with state-of-the-art methods shows the superior performance of the proposed model in structural similarity. Real-world tests confirm its potential for underwater robotics and vision applications.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
This paper presents FreqPatchNet, a novel patch-wise dual-domain Convolutional Neural Network (CNN) designed to correct phase distortions in underwater images. The model uses bispectral frequency features and local CNN regression to reconstruct clean images from distorted inputs. Evaluated using Peak Signal-to-Noise Ratio (PSNR) and Mean Squared Error (MSE), FreqPatchNet achieves a maximum PSNR of 35.6 dB and a lowest MSE of 0.28 at 10% distortion. A comparative analysis with state-of-the-art methods shows the superior performance of the proposed model in structural similarity. Real-world tests confirm its potential for underwater robotics and vision applications.
17.
Sharma, M.; Kumar, Kailash; Gokul, N.; Reddy, G. U. K.; Harshan, N.
Institute of Electrical and Electronics Engineers Inc., 2025, ISBN: 9798331541927 (ISBN), (0).
@proceedings{724,
title = {Transformative Learning Through AI-Driven Handwritten Input: A Computational Tool for Enhanced Mathematical Problem Solving},
author = {M. Sharma and Kailash Kumar and N. Gokul and G. U. K. Reddy and N. Harshan},
url = {https://ieeexplore.ieee.org/document/11210329},
doi = {10.1109/INCSST64791.2025.11210329},
isbn = {9798331541927 (ISBN)},
year = {2025},
date = {2025-01-01},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {This study introduces Transformative Learning Through AI-Driven Handwritten Input, an innovative web-based computational tool designed to enhance mathematical problem-solving for students, educators, and professionals. The application addresses limitations in traditional calculators and digital note-taking tools, which often lack intuitive user experiences and natural input methods.The core objective is to explore whether AI-powered hand-written input can improve the efficiency and accuracy of solving mathematical problems. The tool combines a scientific calculator with a handwriting canvas, offering a more engaging and fluid interface. Developed using React for the front end and Python for the back end, the system processes user-drawn input via a canvas element. The handwritten content is converted into.png images using the Pillow (PIL) library and analyzed through the Gemini API for symbol recognition and computationInitial user testing indicates increased engagement and improved problem-solving speed compared to traditional input methods. Users reported a 30-40% reduction in solution time for complex problems, suggesting that natural handwriting input paired with AI significantly boosts productivity and learning outcomes.This research highlights the potential of integrating AI recognition into educational technology to create more intuitive and effective STEM learning tools. By bridging the gap between analog handwriting and digital computation, this tool introduces an innovative pathway for enhancing digital learning experiences, offering significant value in educational settings such as classrooms, virtual learning platforms, and professional training environments.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
This study introduces Transformative Learning Through AI-Driven Handwritten Input, an innovative web-based computational tool designed to enhance mathematical problem-solving for students, educators, and professionals. The application addresses limitations in traditional calculators and digital note-taking tools, which often lack intuitive user experiences and natural input methods.The core objective is to explore whether AI-powered hand-written input can improve the efficiency and accuracy of solving mathematical problems. The tool combines a scientific calculator with a handwriting canvas, offering a more engaging and fluid interface. Developed using React for the front end and Python for the back end, the system processes user-drawn input via a canvas element. The handwritten content is converted into.png images using the Pillow (PIL) library and analyzed through the Gemini API for symbol recognition and computationInitial user testing indicates increased engagement and improved problem-solving speed compared to traditional input methods. Users reported a 30-40% reduction in solution time for complex problems, suggesting that natural handwriting input paired with AI significantly boosts productivity and learning outcomes.This research highlights the potential of integrating AI recognition into educational technology to create more intuitive and effective STEM learning tools. By bridging the gap between analog handwriting and digital computation, this tool introduces an innovative pathway for enhancing digital learning experiences, offering significant value in educational settings such as classrooms, virtual learning platforms, and professional training environments.
18.
Nachiappan, B.; Najmusher, H.; Nagarajan, G.; Rajkumar, N.; Loganathan, D.
Exploring the Application of Drone Technology in the Construction Sector Journal Article
In: Salud, Ciencia y Tecnologia - Serie de Conferencias, vol. 3, pp. 713+, 2024, ISBN: 29534860 (ISSN), (2).
@article{86,
title = {Exploring the Application of Drone Technology in the Construction Sector},
author = {B. Nachiappan and H. Najmusher and G. Nagarajan and N. Rajkumar and D. Loganathan},
doi = {10.56294/sctconf2024713},
isbn = {29534860 (ISSN)},
year = {2024},
date = {2024-01-01},
journal = {Salud, Ciencia y Tecnologia - Serie de Conferencias},
volume = {3},
pages = {713+},
publisher = {Editorial Salud, Ciencia y Tecnologia},
abstract = {Drone Technology is being used by an increasing number via the development area to improve some of the elements of its operations. Drones have unique competencies that can grow construction projects effectiveness, protection, and affordability. This study examines how drones are presently being used inside the construction zone and the way they could affect construction site online surveying, project control, development monitoring, and safety inspections. The study additionally addresses the difficulties and capacity benefits of incorporating the drone era into production methods. The drone era has a variety of capabilities to change traditional techniques inside the production industry and decorate challenge outcomes. Droneera software creation has become a recreation-changing trend with many benefits, from more suitable productivity to safer and greater sustainable operations. Drones are being used for some purposes, which include environmental tracking, site online surveying, inspections, and development tracking. This study examines the several uses of drones within the construction sector and talks about how this technology may affect the sector going forward. © 2024; Los autores.},
note = {2},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
Drone Technology is being used by an increasing number via the development area to improve some of the elements of its operations. Drones have unique competencies that can grow construction projects effectiveness, protection, and affordability. This study examines how drones are presently being used inside the construction zone and the way they could affect construction site online surveying, project control, development monitoring, and safety inspections. The study additionally addresses the difficulties and capacity benefits of incorporating the drone era into production methods. The drone era has a variety of capabilities to change traditional techniques inside the production industry and decorate challenge outcomes. Droneera software creation has become a recreation-changing trend with many benefits, from more suitable productivity to safer and greater sustainable operations. Drones are being used for some purposes, which include environmental tracking, site online surveying, inspections, and development tracking. This study examines the several uses of drones within the construction sector and talks about how this technology may affect the sector going forward. © 2024; Los autores.
19.
Madhumohan, A. J. V.; Raghu, H.; Priyanka, R.; Dsouza, S. M.; Teena, K. B.; Shetty, S.
Automated Text Extraction and Classification from Images using Multi-Layer Perceptron(MLP) Proceedings
Institute of Electrical and Electronics Engineers Inc., 2024, ISBN: 979-835035968-8 (ISBN), (0).
@proceedings{92,
title = {Automated Text Extraction and Classification from Images using Multi-Layer Perceptron(MLP)},
author = {A. J. V. Madhumohan and H. Raghu and R. Priyanka and S. M. Dsouza and K. B. Teena and S. Shetty},
doi = {10.1109/ICKECS61492.2024.10617274},
isbn = {979-835035968-8 (ISBN)},
year = {2024},
date = {2024-01-01},
journal = {2024 International Conference on Knowledge Engineering and Communication Systems, ICKECS 2024},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {This research paper presents a comprehensive approach for extracting and classifying text from images using computer vision and deep learning techniques. We demonstrate a step-by-step process of preprocessing, text extraction, and classification with a Multi-Layer Perceptron(MLP). The proposed method is evaluated on a dataset of player images having different kill counts and shows promising results using bounding box algorithm in automated text extraction and classification. © 2024 IEEE.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
This research paper presents a comprehensive approach for extracting and classifying text from images using computer vision and deep learning techniques. We demonstrate a step-by-step process of preprocessing, text extraction, and classification with a Multi-Layer Perceptron(MLP). The proposed method is evaluated on a dataset of player images having different kill counts and shows promising results using bounding box algorithm in automated text extraction and classification. © 2024 IEEE.
20.
Priyanka, D.
Educational technology and libraries supporting online/digital learning with the ASP.NET MVC framework Book Chapter
In: pp. 191-208,, IGI Global, 2024, ISBN: 979-836932783-8 (ISBN); 979-836932782-1 (ISBN), (2).
@inbook{67,
title = {Educational technology and libraries supporting online/digital learning with the ASP.NET MVC framework},
author = {D. Priyanka},
doi = {10.4018/979-8-3693-2782-1.ch011},
isbn = {979-836932783-8 (ISBN); 979-836932782-1 (ISBN)},
year = {2024},
date = {2024-01-01},
journal = {AI-Assisted Library Reconstruction},
pages = {191-208,},
publisher = {IGI Global},
abstract = {Online applications are quickly multiplying across different areas, encompassing large and small enterprises, government entities, academic institutions, and research centers. Utilizing the MVC (modelview-regulator) philosophy, this chapter advocates for a clear separation of concerns, offering numerous advantages. Existing writing highlights an absence or a lack of a unified approach within ASP.NET MVC to effectively address identified challenges, hindering the optimization of online learning experiences in educational settings. This proposed arrangement, utilizing the qualities of ASP.NET MVC and consolidating man-made intelligence help, focuses on particular advancement for the making of easy- to-understand interfaces. The chapter highlights the significant advantages emerging from this organization system in the domain of instructive innovation and libraries. Through a detailed exploration of this approach, the authors aim to significantly contribute to the advancement of AI-assisted educational technology and libraries. © 2024, IGI Global. All rights reserved.},
note = {2},
keywords = {ISE},
pubstate = {published},
tppubtype = {inbook}
}
Online applications are quickly multiplying across different areas, encompassing large and small enterprises, government entities, academic institutions, and research centers. Utilizing the MVC (modelview-regulator) philosophy, this chapter advocates for a clear separation of concerns, offering numerous advantages. Existing writing highlights an absence or a lack of a unified approach within ASP.NET MVC to effectively address identified challenges, hindering the optimization of online learning experiences in educational settings. This proposed arrangement, utilizing the qualities of ASP.NET MVC and consolidating man-made intelligence help, focuses on particular advancement for the making of easy- to-understand interfaces. The chapter highlights the significant advantages emerging from this organization system in the domain of instructive innovation and libraries. Through a detailed exploration of this approach, the authors aim to significantly contribute to the advancement of AI-assisted educational technology and libraries. © 2024, IGI Global. All rights reserved.
21.
Vathsala, M. K.; Lingareddy, S. C.
Transliteration and translation of the Hindi language using integrated domain-based auto-encoder Journal Article
In: IAES International Journal of Artificial Intelligence, vol. 13, pp. 4906-4914,, 2024, ISBN: 20894872 (ISSN), (0).
@article{54,
title = {Transliteration and translation of the Hindi language using integrated domain-based auto-encoder},
author = {M. K. Vathsala and S. C. Lingareddy},
doi = {10.11591/ijai.v13.i4.pp4906-4914},
isbn = {20894872 (ISSN)},
year = {2024},
date = {2024-01-01},
journal = {IAES International Journal of Artificial Intelligence},
volume = {13},
pages = {4906-4914,},
publisher = {Institute of Advanced Engineering and Science},
abstract = {The main objective of translation is to translate words' meanings from one language to another; in contrast, transliteration does not translate any contextual meanings between languages. Transliteration, as opposed to translation, just considers the individual letters that make up each word. In this paper, an integrated deep neural network transliteration and translation model (NNTT) based autoencoder model is developed. The model is segmented into transliteration model and translation model; the transliteration involves the process of converting text from one script to another evaluated on the Dakshina dataset wherein Hindi typically uses a sequence-to-sequence model with an attention mechanism, the translation model is trained to translate text from one language to another. Translation models regularly use a sequence-to-sequence model performed on the workshop on Asian translation (WAT) 2021 dataset with an attention mechanism, similar to the one used in the transliteration model for Hindi. The proposed NNTT model merges the in-domain and out-domain frameworks to develop a training framework so that the information is transferred between the domains. The results evaluated show that the proposed model works effectively in comparison with the existing system for the Hindi language. © 2024, Institute of Advanced Engineering and Science. All rights reserved.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
The main objective of translation is to translate words' meanings from one language to another; in contrast, transliteration does not translate any contextual meanings between languages. Transliteration, as opposed to translation, just considers the individual letters that make up each word. In this paper, an integrated deep neural network transliteration and translation model (NNTT) based autoencoder model is developed. The model is segmented into transliteration model and translation model; the transliteration involves the process of converting text from one script to another evaluated on the Dakshina dataset wherein Hindi typically uses a sequence-to-sequence model with an attention mechanism, the translation model is trained to translate text from one language to another. Translation models regularly use a sequence-to-sequence model performed on the workshop on Asian translation (WAT) 2021 dataset with an attention mechanism, similar to the one used in the transliteration model for Hindi. The proposed NNTT model merges the in-domain and out-domain frameworks to develop a training framework so that the information is transferred between the domains. The results evaluated show that the proposed model works effectively in comparison with the existing system for the Hindi language. © 2024, Institute of Advanced Engineering and Science. All rights reserved.
22.
Priyanka, R.; Teena, K. B.; Rashmi, T. V.; Reshma, J.; Nagaraj, T.; Tejaswini, N.
A Hybrid Cluster Based Intelligent IDS with Deep Belief Network to Improve the Security over Wireless Sensor Network Journal Article
In: International Journal of Intelligent Systems and Applications in Engineering, vol. 12, pp. 225-238,, 2024, ISBN: 21476799 (ISSN), (0).
@article{34,
title = {A Hybrid Cluster Based Intelligent IDS with Deep Belief Network to Improve the Security over Wireless Sensor Network},
author = {R. Priyanka and K. B. Teena and T. V. Rashmi and J. Reshma and T. Nagaraj and N. Tejaswini},
isbn = {21476799 (ISSN)},
year = {2024},
date = {2024-01-01},
journal = {International Journal of Intelligent Systems and Applications in Engineering},
volume = {12},
pages = {225-238,},
publisher = {Ismail Saritas},
abstract = {Numerous inexpensive, compact devices compose a Wireless Sensor Network (WSN). They're usually readily available to some types of attacks due to their location, which is not well protected. A large number of researchers are focusing on WSN security at the moment. This kind of network is characterized by vulnerable characteristics, such as the ability to organize oneself without a stable infrastructure and open-air transmission. To train variables for the probability-based feature vectors, a Deep Neural Network (DNN) framework that is derived from international vehicle network packets shall be applied. The detector is capable of detecting any malicious attack on the vehicle since DNN gives each category a chance to distinguish between attacks and regular packets. Intrusion Detection Systems (IDS), can help to identify and stop security attacks on vehicles. The study proposes a mechanism for enhancing the security of WSNs based on Hybrid Clusters and Intelligent Intrusion Detection Systems with Deep Belief Networks (HCIIDS-DBN). It can provide a protection system for intrusions and an analysis of vehicle attacks in real time. They are designed based on their respective attack probability and ability, to the sensor node, sink, or cluster head. The proposed HCIIDS-DBN is composed of modules designed to detect anomalies and dereliction. The objective is to increase detection rates and decrease false positive incidences by detecting anomalies and abuse. Finally, the detected data are integrated and the various types of vehicle communication attacks are reported using the Decision Support System (DSS). The results of the experiment show that the proposed method may respond to the attack in real-time with a much detection of higher ratio in the Controller Area Network (CAN) bus. © 2024, Ismail Saritas. All rights reserved.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
Numerous inexpensive, compact devices compose a Wireless Sensor Network (WSN). They're usually readily available to some types of attacks due to their location, which is not well protected. A large number of researchers are focusing on WSN security at the moment. This kind of network is characterized by vulnerable characteristics, such as the ability to organize oneself without a stable infrastructure and open-air transmission. To train variables for the probability-based feature vectors, a Deep Neural Network (DNN) framework that is derived from international vehicle network packets shall be applied. The detector is capable of detecting any malicious attack on the vehicle since DNN gives each category a chance to distinguish between attacks and regular packets. Intrusion Detection Systems (IDS), can help to identify and stop security attacks on vehicles. The study proposes a mechanism for enhancing the security of WSNs based on Hybrid Clusters and Intelligent Intrusion Detection Systems with Deep Belief Networks (HCIIDS-DBN). It can provide a protection system for intrusions and an analysis of vehicle attacks in real time. They are designed based on their respective attack probability and ability, to the sensor node, sink, or cluster head. The proposed HCIIDS-DBN is composed of modules designed to detect anomalies and dereliction. The objective is to increase detection rates and decrease false positive incidences by detecting anomalies and abuse. Finally, the detected data are integrated and the various types of vehicle communication attacks are reported using the Decision Support System (DSS). The results of the experiment show that the proposed method may respond to the attack in real-time with a much detection of higher ratio in the Controller Area Network (CAN) bus. © 2024, Ismail Saritas. All rights reserved.
23.
Salagare, S.; Sudha, P. N.; Palani, K.
Sustainable energy harvesting system for low-power underwater sensing devices Journal Article
In: Indonesian Journal of Electrical Engineering and Computer Science, vol. 35, pp. 1379-1387,, 2024, ISBN: 25024752 (ISSN), (0).
@article{32,
title = {Sustainable energy harvesting system for low-power underwater sensing devices},
author = {S. Salagare and P. N. Sudha and K. Palani},
doi = {10.11591/ijeecs.v35.i3.pp1379-1387},
isbn = {25024752 (ISSN)},
year = {2024},
date = {2024-01-01},
journal = {Indonesian Journal of Electrical Engineering and Computer Science},
volume = {35},
pages = {1379-1387,},
publisher = {Institute of Advanced Engineering and Science},
abstract = {In marine scientific research, ocean monitoring is crucial where the battery-powered sensor devices are placed under the water to collect different information like temperature, pressure, and turbidity in underwater sensor networks (UWSNs). Thus, keeping these devices active for longer periods is challenging. In the last decades, the piezoelectric transducer (PZT) material has been used widely for constructing more environmentally friendly energy harvesting systems. The PZT harvester offers a promising solution by eliminating the need for batteries for running devices in the future with less maintenance. The PZT harvester allows the system to generate higher voltage to run low-power devices. This paper designed and developed a new renewable energy harvester system using PZT transducers for running different types of underwater sensor devices like temperature, turbidity, and obstacle sensors. The proposed PZT-based energy harvester employs a two-stage amplification model for generating higher voltage and current to run multiple devices. The sensing information collected from these sensors is transmitted to the cloud which is later utilized for analysis and decision-making. Experiment results show the proposed PZT-based energy harvester can generate a voltage of 13 volts (V) and a current of 43.3 milliampere (mA) equivalent to 562 milliwatt (mW) which is very good to run multiple low-power underwater sensor devices. © 2024 Institute of Advanced Engineering and Science. All rights reserved.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
In marine scientific research, ocean monitoring is crucial where the battery-powered sensor devices are placed under the water to collect different information like temperature, pressure, and turbidity in underwater sensor networks (UWSNs). Thus, keeping these devices active for longer periods is challenging. In the last decades, the piezoelectric transducer (PZT) material has been used widely for constructing more environmentally friendly energy harvesting systems. The PZT harvester offers a promising solution by eliminating the need for batteries for running devices in the future with less maintenance. The PZT harvester allows the system to generate higher voltage to run low-power devices. This paper designed and developed a new renewable energy harvester system using PZT transducers for running different types of underwater sensor devices like temperature, turbidity, and obstacle sensors. The proposed PZT-based energy harvester employs a two-stage amplification model for generating higher voltage and current to run multiple devices. The sensing information collected from these sensors is transmitted to the cloud which is later utilized for analysis and decision-making. Experiment results show the proposed PZT-based energy harvester can generate a voltage of 13 volts (V) and a current of 43.3 milliampere (mA) equivalent to 562 milliwatt (mW) which is very good to run multiple low-power underwater sensor devices. © 2024 Institute of Advanced Engineering and Science. All rights reserved.
24.
Shankar, M. S.; Adishesha, R.; Kumar, Hemanth; Jayanthi, M. G.; Kannadaguli, P.; Loganathan, D.
Image-Based Plant Disease Classification for the Management of Crop Health Proceedings
Institute of Electrical and Electronics Engineers Inc., 2024, ISBN: 979-835034367-0 (ISBN), (1).
@proceedings{27,
title = {Image-Based Plant Disease Classification for the Management of Crop Health},
author = {M. S. Shankar and R. Adishesha and Hemanth Kumar and M. G. Jayanthi and P. Kannadaguli and D. Loganathan},
doi = {10.1109/ICAECT60202.2024.10469390},
isbn = {979-835034367-0 (ISBN)},
year = {2024},
date = {2024-01-01},
journal = {2024 4th International Conference on Advances in Electrical, Computing, Communication and Sustainable Technologies, ICAECT 2024},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {Categorizing plant diseases is crucial for ensuring agricultural production and food security. In this research, we investigate two distinct methods for classifying plant diseases: Convolutional neural networks (CNN) for deep learning and logistic regression (LR) along with Random Forest Classifier (RFC) for machine learning. We use a collection of plant pictures representing various diseases to train and evaluate LR and CNN models. The CNN model automatically learns hierarchical representations, while the LR model relies on manually created features extracted from the images. Our analysis reveals that both LR and CNN models achieve high accuracy in classifying plant diseases, with CNN surpassing LR due to its ability to recognize complex image patterns. The CNN model's performance in our experiment outperforms other models in terms of accuracy. The experiment's findings underscore the effectiveness of deep learning and machine learning techniques in classifying plant diseases. © 2024 IEEE.},
note = {1},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
Categorizing plant diseases is crucial for ensuring agricultural production and food security. In this research, we investigate two distinct methods for classifying plant diseases: Convolutional neural networks (CNN) for deep learning and logistic regression (LR) along with Random Forest Classifier (RFC) for machine learning. We use a collection of plant pictures representing various diseases to train and evaluate LR and CNN models. The CNN model automatically learns hierarchical representations, while the LR model relies on manually created features extracted from the images. Our analysis reveals that both LR and CNN models achieve high accuracy in classifying plant diseases, with CNN surpassing LR due to its ability to recognize complex image patterns. The CNN model's performance in our experiment outperforms other models in terms of accuracy. The experiment's findings underscore the effectiveness of deep learning and machine learning techniques in classifying plant diseases. © 2024 IEEE.
25.
V, S.; Abdullah, A.; Ramadass, P.; Srinivasan, S.; Shivahare, B. D.; Mathivanan, S. K.; P, K.
Context based ranking strategies for renowned instructional methodologies Journal Article
In: Intelligence-Based Medicine, vol. 10, pp. 100186+, 2024, ISBN: 26665212 (ISSN), (0).
@article{20,
title = {Context based ranking strategies for renowned instructional methodologies},
author = {S. V and A. Abdullah and P. Ramadass and S. Srinivasan and B. D. Shivahare and S. K. Mathivanan and K. P},
doi = {10.1016/j.ibmed.2024.100186},
isbn = {26665212 (ISSN)},
year = {2024},
date = {2024-01-01},
journal = {Intelligence-Based Medicine},
volume = {10},
pages = {100186+},
publisher = {Elsevier B.V.},
abstract = {The main objective of this work is to validate the decisions made towards adoption of appropriate instructional methodologies based on the context of a specific region considering the quality of education, the cost of education and the learning outcomes as predominant parameters. The non-deterministic events and uncertain situations that may arise over a long-range period impose a vague and fuzzy environment in the educational system. Investigations have been made to identify suitable educational framework for implementation in the institutions of a specific region in view of these unpredictable events and non-deterministic conditions. Fuzzy decision analysis and rough set theory have been applied to rank the prominent instructional methodologies which are encompassed within each educational framework. Hurwicz Rule is adopted to balance the pessimistic and optimistic opinions about the non-deterministic events while validating the merits of the instructional methodologies. Grey relational analysis is carried out while ranking instructional methodologies in a vague environment. In this work, the instructional methodologies are ranked using fuzzy entropy as well as crisp entropy measures and the outcomes of the fuzzy and rough sets-based decision analysis have been validated. © 2024 The Authors},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
The main objective of this work is to validate the decisions made towards adoption of appropriate instructional methodologies based on the context of a specific region considering the quality of education, the cost of education and the learning outcomes as predominant parameters. The non-deterministic events and uncertain situations that may arise over a long-range period impose a vague and fuzzy environment in the educational system. Investigations have been made to identify suitable educational framework for implementation in the institutions of a specific region in view of these unpredictable events and non-deterministic conditions. Fuzzy decision analysis and rough set theory have been applied to rank the prominent instructional methodologies which are encompassed within each educational framework. Hurwicz Rule is adopted to balance the pessimistic and optimistic opinions about the non-deterministic events while validating the merits of the instructional methodologies. Grey relational analysis is carried out while ranking instructional methodologies in a vague environment. In this work, the instructional methodologies are ranked using fuzzy entropy as well as crisp entropy measures and the outcomes of the fuzzy and rough sets-based decision analysis have been validated. © 2024 The Authors
26.
Gayathri, T.; Mahalakshmi, K.; Shilpa, M.; Jayanthi, M. G.; Kannadaguli, P.
Comparison of Hate Speech Identification in Kannada Language Using ML and DL Models Proceedings
Institute of Electrical and Electronics Engineers Inc., 2023, ISBN: 979-835030816-7 (ISBN), (0).
@proceedings{14,
title = {Comparison of Hate Speech Identification in Kannada Language Using ML and DL Models},
author = {T. Gayathri and K. Mahalakshmi and M. Shilpa and M. G. Jayanthi and P. Kannadaguli},
doi = {10.1109/GCITC60406.2023.10425987},
isbn = {979-835030816-7 (ISBN)},
year = {2023},
date = {2023-01-01},
journal = {2023 Global Conference on Information Technologies and Communications, GCITC 2023},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {The problem at hand is to create a discrimination system specifically for Indian languages, with an emphasis on Automatic Speech Recognition (ASR) implementations. Hate speech poses a serious challenge to online websites and social media, as well as causing harm, such as spreading hate, inciting violence, and promoting inequality. Macro skills are discriminatory, there is an urgent need for a similar system for regional languages as the country has many different languages and unique cultures. Therefore, this paper intends to gauge the overall performance of 4 characteristic engineering strategies and 4 gadget learning algorithms to examine their overall performance on a publicly-to-be-had dataset with two distinct classes. The experimental consequences confirmed that the bigram capabilities when used with the help vector machine set of rules great carried out with 88% accuracy in ML and 91% of accuracy in DL. This observation has practical implications and can be used as a basis for detecting automated hate speech messages. Moreover, the output of different affinity could be utilized as country-of-artwork strategies to compare destiny research for existing computerized text classification techniques. © 2023 IEEE.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
The problem at hand is to create a discrimination system specifically for Indian languages, with an emphasis on Automatic Speech Recognition (ASR) implementations. Hate speech poses a serious challenge to online websites and social media, as well as causing harm, such as spreading hate, inciting violence, and promoting inequality. Macro skills are discriminatory, there is an urgent need for a similar system for regional languages as the country has many different languages and unique cultures. Therefore, this paper intends to gauge the overall performance of 4 characteristic engineering strategies and 4 gadget learning algorithms to examine their overall performance on a publicly-to-be-had dataset with two distinct classes. The experimental consequences confirmed that the bigram capabilities when used with the help vector machine set of rules great carried out with 88% accuracy in ML and 91% of accuracy in DL. This observation has practical implications and can be used as a basis for detecting automated hate speech messages. Moreover, the output of different affinity could be utilized as country-of-artwork strategies to compare destiny research for existing computerized text classification techniques. © 2023 IEEE.
27.
Sharma, M.; Supriya, M.; Kumar, A.; Dhyani, K.; Chaturvedi, P.
Extraction of Water and Riverine Sand using Deep Learning on Multispectral Remote Sensing Images Proceedings
Institute of Electrical and Electronics Engineers Inc., 2023, ISBN: 979-835034060-0 (ISBN), (0).
@proceedings{8,
title = {Extraction of Water and Riverine Sand using Deep Learning on Multispectral Remote Sensing Images},
author = {M. Sharma and M. Supriya and A. Kumar and K. Dhyani and P. Chaturvedi},
doi = {10.1109/ICECA58529.2023.10395559},
isbn = {979-835034060-0 (ISBN)},
year = {2023},
date = {2023-01-01},
journal = {7th International Conference on Electronics, Communication and Aerospace Technology, ICECA 2023 - Proceedings},
pages = {849-854,},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {The study of land-use-land cover (LULC) has become a necessity with the advancement of the urbanization process. Increased erosion of soil, increased silting, and sedimentation of the rivers are key effects that require study and analysis. Deep learning has a significant impact on classification tasks, particularly in the field of remote sensing image analysis. The proposed framework classifies LULC classes by employing the characteristics of deep learning. In this work, we compared the proposed method with the traditional machine learning methods in extracting water and riverine sand from multispectral remote sensing images. Further, we analyse the impact of Stochastic Gradient Descent (SGD) and Adam optimizers. The Adam optimizer implemented in this work gives higher accuracy than other combinations. © 2023 IEEE.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
The study of land-use-land cover (LULC) has become a necessity with the advancement of the urbanization process. Increased erosion of soil, increased silting, and sedimentation of the rivers are key effects that require study and analysis. Deep learning has a significant impact on classification tasks, particularly in the field of remote sensing image analysis. The proposed framework classifies LULC classes by employing the characteristics of deep learning. In this work, we compared the proposed method with the traditional machine learning methods in extracting water and riverine sand from multispectral remote sensing images. Further, we analyse the impact of Stochastic Gradient Descent (SGD) and Adam optimizers. The Adam optimizer implemented in this work gives higher accuracy than other combinations. © 2023 IEEE.
28.
Desai, P.; Preethi, S.; Loganathan, D.; Bharani, B. R.
Institute of Electrical and Electronics Engineers Inc., 2023, ISBN: 979-835034279-6 (ISBN), (0).
@proceedings{25,
title = {Qualitative and Quantitative Data Analysis using Classification, and Ensemble Techniques to Optimize and Predict the Performance of Reviews},
author = {P. Desai and S. Preethi and D. Loganathan and B. R. Bharani},
doi = {10.1109/ICAEECI58247.2023.10370889},
isbn = {979-835034279-6 (ISBN)},
year = {2023},
date = {2023-01-01},
journal = {2023 1st International Conference on Advances in Electrical, Electronics and Computational Intelligence, ICAEECI 2023},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {As an analyst deeply engaged in data analysis, it becomes imperative to discern the inherent nature of the data, classifying it into either qualitative or quantitative forms. Qualitative data necessitates preprocessing to facilitate predictive modeling of its outcomes. In the context of this research, we employ a movie review dataset to predict both negative and positive reviews. The realm of qualitative analysis faces a notable challenge in predictive capabilities, primarily due to the diverse sentiments expressed in various reviews. To address this challenge, we employ a diverse array of classifiers such as bagging, boosting and stacking to evaluate their performance in terms of accuracy, F1 score, and training time by selecting the best performers as an ensemble classifier. Subsequently, we identify the most effective classifier and apply ensemble techniques and stacking methodologies to optimize predictive accuracy. © 2023 IEEE.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
As an analyst deeply engaged in data analysis, it becomes imperative to discern the inherent nature of the data, classifying it into either qualitative or quantitative forms. Qualitative data necessitates preprocessing to facilitate predictive modeling of its outcomes. In the context of this research, we employ a movie review dataset to predict both negative and positive reviews. The realm of qualitative analysis faces a notable challenge in predictive capabilities, primarily due to the diverse sentiments expressed in various reviews. To address this challenge, we employ a diverse array of classifiers such as bagging, boosting and stacking to evaluate their performance in terms of accuracy, F1 score, and training time by selecting the best performers as an ensemble classifier. Subsequently, we identify the most effective classifier and apply ensemble techniques and stacking methodologies to optimize predictive accuracy. © 2023 IEEE.
29.
Sapna, G. S.; Revanna, S. D.
An Efficient Internet of Things Interoperability Model Using Secure Access Control Mechanism Journal Article
In: International Journal of Intelligent Engineering and Systems, vol. 16, pp. 41-56,, 2023, ISBN: 2185310X (ISSN), (0).
@article{47,
title = {An Efficient Internet of Things Interoperability Model Using Secure Access Control Mechanism},
author = {G. S. Sapna and S. D. Revanna},
doi = {10.22266/ijies2023.1031.05},
isbn = {2185310X (ISSN)},
year = {2023},
date = {2023-01-01},
journal = {International Journal of Intelligent Engineering and Systems},
volume = {16},
pages = {41-56,},
publisher = {Intelligent Network and Systems Society},
abstract = {Internet of Things (IoT) is a revolutionary innovation in many aspects of our society like financial activities, communication activities, and global security such as the military and battlefields’ internet. Security and energy play a crucial role in data transmission across IoT and edge networks. In this research, a trust mechanism based on privacy access control is proposed for IoT devices’ interoperability. Most of the existing researches on achieving interoperability for IoT devices has drawbacks such as overlapping of systems, uneven distribution of data, lack of data security, high power consumption, and low optimization of resources. The main objective of this research is to focus and overcome these challenges by introducing a privacy access control mechanism that includes trust parameters of IoT device interoperability. A routing protocol for low-power and lossy networks (RPL) mode of operation is set in the direction of multipoint-to-point traffic flow, except in the downward flow direction. Sensor nodes send data packets to the sink node, which transmits the information to the server to determine the trust values in this mode. To validate the performance, a widely used lightweight low-power wireless simulator Contiki/cooja simulator is implemented. The simulation results of the proposed model have shown a transmission ratio of 100%, a receiver ratio of 30 to 100%, and the detection of malicious nodes in a simulation time of 60 minutes. With the use of the proposed trust mechanism based on privacy access control, a less packet loss ratio of 0.43% is achieved along with less power consumption of 0.4%, and the highest average residual energy of 0.87mJoules at node 30. © (2023), (Intelligent Network and Systems Society). All Rights Reserved.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
Internet of Things (IoT) is a revolutionary innovation in many aspects of our society like financial activities, communication activities, and global security such as the military and battlefields’ internet. Security and energy play a crucial role in data transmission across IoT and edge networks. In this research, a trust mechanism based on privacy access control is proposed for IoT devices’ interoperability. Most of the existing researches on achieving interoperability for IoT devices has drawbacks such as overlapping of systems, uneven distribution of data, lack of data security, high power consumption, and low optimization of resources. The main objective of this research is to focus and overcome these challenges by introducing a privacy access control mechanism that includes trust parameters of IoT device interoperability. A routing protocol for low-power and lossy networks (RPL) mode of operation is set in the direction of multipoint-to-point traffic flow, except in the downward flow direction. Sensor nodes send data packets to the sink node, which transmits the information to the server to determine the trust values in this mode. To validate the performance, a widely used lightweight low-power wireless simulator Contiki/cooja simulator is implemented. The simulation results of the proposed model have shown a transmission ratio of 100%, a receiver ratio of 30 to 100%, and the detection of malicious nodes in a simulation time of 60 minutes. With the use of the proposed trust mechanism based on privacy access control, a less packet loss ratio of 0.43% is achieved along with less power consumption of 0.4%, and the highest average residual energy of 0.87mJoules at node 30. © (2023), (Intelligent Network and Systems Society). All Rights Reserved.
30.
Soubhagyalakshmi, P.; Reddy, K. S.
An efficient security analysis of bring your own device Journal Article
In: IAES International Journal of Artificial Intelligence, vol. 12, pp. 696-703,, 2023, ISBN: 20894872 (ISSN), (3).
@article{62,
title = {An efficient security analysis of bring your own device},
author = {P. Soubhagyalakshmi and K. S. Reddy},
doi = {10.11591/ijai.v12.i2.pp696-703},
isbn = {20894872 (ISSN)},
year = {2023},
date = {2023-01-01},
journal = {IAES International Journal of Artificial Intelligence},
volume = {12},
pages = {696-703,},
publisher = {Institute of Advanced Engineering and Science},
abstract = {The significant enhancement in demand for bring your own device (BYOD) mechanism in several organizations has sought the attention of several researchers in recent years. However, the utilization of BYOD comes with a high risk of losing crucial information due to lesser organizational control on employee-owned devices. The purpose of this article is to review and analyze the various security threats in BYOD; further we review the existing work that was developed in order to reduce the risks present in BYOD. A detailed review is presented to detect BYOD security threats and their respective security policies. A phase-by-phase mitigation strategy is developed based on the components and crucial elements identified using review policy. Managerial-level, social-level and technical level issues are identified such as illegal access, leaking delicate company data, lower flexibility, corporate data breaching, and employee privacy. It is analyzed that collaboration of people, security policy factors and technology in an effective manner can mitigate security threats present in the BYOD mechanism. This article initiates a move towards filling the security gap present the BYOD mechanism. This article can be utilized for providing guidelines in various organizations. Ultimately, successful implementation of BYOD depends upon the balance created between usability and security. © 2023, Institute of Advanced Engineering and Science. All rights reserved.},
note = {3},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
The significant enhancement in demand for bring your own device (BYOD) mechanism in several organizations has sought the attention of several researchers in recent years. However, the utilization of BYOD comes with a high risk of losing crucial information due to lesser organizational control on employee-owned devices. The purpose of this article is to review and analyze the various security threats in BYOD; further we review the existing work that was developed in order to reduce the risks present in BYOD. A detailed review is presented to detect BYOD security threats and their respective security policies. A phase-by-phase mitigation strategy is developed based on the components and crucial elements identified using review policy. Managerial-level, social-level and technical level issues are identified such as illegal access, leaking delicate company data, lower flexibility, corporate data breaching, and employee privacy. It is analyzed that collaboration of people, security policy factors and technology in an effective manner can mitigate security threats present in the BYOD mechanism. This article initiates a move towards filling the security gap present the BYOD mechanism. This article can be utilized for providing guidelines in various organizations. Ultimately, successful implementation of BYOD depends upon the balance created between usability and security. © 2023, Institute of Advanced Engineering and Science. All rights reserved.
31.
Acharya, N.; Singh, A. K.; Dwivedi, A. K.; Jayanthi, M. G.; Kannadaguli, P.
Indian Food Segmentation and Calorie Estimation Using CatBoost and Masked Convolutional Neural Networks Proceedings
Institute of Electrical and Electronics Engineers Inc., 2023, ISBN: 979-835030082-6 (ISBN), (1).
@proceedings{64,
title = {Indian Food Segmentation and Calorie Estimation Using CatBoost and Masked Convolutional Neural Networks},
author = {N. Acharya and A. K. Singh and A. K. Dwivedi and M. G. Jayanthi and P. Kannadaguli},
doi = {10.1109/NMITCON58196.2023.10275885},
isbn = {979-835030082-6 (ISBN)},
year = {2023},
date = {2023-01-01},
journal = {2023 International Conference on Network, Multimedia and Information Technology, NMITCON 2023},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {Accurate food segmentation and calorie estimation play a pivotal role in effective dietary monitoring and health management. In this paper, after extensive experimentation on a diverse Indian food dataset of 20 classes collected by us, we present two separate innovative models designed to address these tasks with exceptional precision. The first model employs a machine learning approach by incorporating the CatBoost algorithm, while the second model leverages a deep learning technique utilizing the U-Net architecture for high-quality image segmentation. Additionally, we merge the U-Net outcomes with a convolutional neural network (CNN) to enhance the deep learning-based classification. Our models effectively handle categorical features and address imbalanced data, resulting in significantly improved accuracy in food item delineation within images. This advancement enables a more reliable and comprehensive analysis of dietary patterns. We demonstrate the effectiveness and robustness of our proposed models. Comparative evaluations against state-of-the-art methods verify exceptional performance in terms of segmentation accuracy, precision in food classification, and estimation of calorie intake. Notably, our models are specifically tailored for food images and do not take ingredients or any other information into consideration. In comparison CNN model is performs better than CatBoost, achieving an accuracy of 78%. © 2023 IEEE.},
note = {1},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
Accurate food segmentation and calorie estimation play a pivotal role in effective dietary monitoring and health management. In this paper, after extensive experimentation on a diverse Indian food dataset of 20 classes collected by us, we present two separate innovative models designed to address these tasks with exceptional precision. The first model employs a machine learning approach by incorporating the CatBoost algorithm, while the second model leverages a deep learning technique utilizing the U-Net architecture for high-quality image segmentation. Additionally, we merge the U-Net outcomes with a convolutional neural network (CNN) to enhance the deep learning-based classification. Our models effectively handle categorical features and address imbalanced data, resulting in significantly improved accuracy in food item delineation within images. This advancement enables a more reliable and comprehensive analysis of dietary patterns. We demonstrate the effectiveness and robustness of our proposed models. Comparative evaluations against state-of-the-art methods verify exceptional performance in terms of segmentation accuracy, precision in food classification, and estimation of calorie intake. Notably, our models are specifically tailored for food images and do not take ingredients or any other information into consideration. In comparison CNN model is performs better than CatBoost, achieving an accuracy of 78%. © 2023 IEEE.
32.
Sapna, G. S.; Revanna, S. D.
An Interoperability Framework for Enhanced Security of Handheld Devices Using IoT-Based Secure Energy Efficient Firefly Optimization Algorithm Journal Article
In: International Journal of Computer Networks and Applications, vol. 10, pp. 763-775,, 2023, ISBN: 23950455 (ISSN), (0).
@article{78,
title = {An Interoperability Framework for Enhanced Security of Handheld Devices Using IoT-Based Secure Energy Efficient Firefly Optimization Algorithm},
author = {G. S. Sapna and S. D. Revanna},
doi = {10.22247/ijcna/2023/223422},
isbn = {23950455 (ISSN)},
year = {2023},
date = {2023-01-01},
journal = {International Journal of Computer Networks and Applications},
volume = {10},
pages = {763-775,},
publisher = {EverScience Publications},
abstract = {Security is a major challenge in the Internet of Things (IoT) domain as it plays a crucial role in a safe and uninterrupted data transmission, across various hand-held devices connected to the network. Establishing a secure Routing Protocol for Low power and lossy networks (RPL) is necessary and crucial, as it is the standard RPL network in IoT that helps to remove malicious nodes from the network. The existing researches focused on developing energy-saving techniques, malicious node detection techniques, as well as security-enhancing techniques, but neglected energy efficiency, and other trust-related considerations. This resulted in reduced confidentiality and unauthorized access to user data. To overcome these limitations, a Secure Energy Efficient Firefly Optimization Algorithm in RPL (SEEFOA-RPL) is proposed in this research for establishing a reliable and energy-efficient routing path by using Destination-Oriented Directed Acyclic Graph (DODAG) architecture. The proposed algorithm improves security measures in handheld devices such as smartphones, wearable watches, digital cameras, portable media players, and tablets. Initially, a trust model for the RPL network is established to calculate the trust parameters that help in building a secure routing in the network. The SEEFOA is capable of solving complex optimization problems, and finds the best optimum solution for a secure-energy efficient routing path. The proposed SEEFOA-RPL delivers a high-level performance in terms of Detection Rate (DR), False Negative Rate (FNR), and False Positive Rate (FPR), respectively measured at 99%, 12%, and 17% in an attack interval 4, and Packet Drop Ratio (PDR) measured at 82% in an attack interval of 1.5. © 2023 EverScience Publications. All rights reserved},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
Security is a major challenge in the Internet of Things (IoT) domain as it plays a crucial role in a safe and uninterrupted data transmission, across various hand-held devices connected to the network. Establishing a secure Routing Protocol for Low power and lossy networks (RPL) is necessary and crucial, as it is the standard RPL network in IoT that helps to remove malicious nodes from the network. The existing researches focused on developing energy-saving techniques, malicious node detection techniques, as well as security-enhancing techniques, but neglected energy efficiency, and other trust-related considerations. This resulted in reduced confidentiality and unauthorized access to user data. To overcome these limitations, a Secure Energy Efficient Firefly Optimization Algorithm in RPL (SEEFOA-RPL) is proposed in this research for establishing a reliable and energy-efficient routing path by using Destination-Oriented Directed Acyclic Graph (DODAG) architecture. The proposed algorithm improves security measures in handheld devices such as smartphones, wearable watches, digital cameras, portable media players, and tablets. Initially, a trust model for the RPL network is established to calculate the trust parameters that help in building a secure routing in the network. The SEEFOA is capable of solving complex optimization problems, and finds the best optimum solution for a secure-energy efficient routing path. The proposed SEEFOA-RPL delivers a high-level performance in terms of Detection Rate (DR), False Negative Rate (FNR), and False Positive Rate (FPR), respectively measured at 99%, 12%, and 17% in an attack interval 4, and Packet Drop Ratio (PDR) measured at 82% in an attack interval of 1.5. © 2023 EverScience Publications. All rights reserved
33.
Ravi, J.; Rajkumar, N.; Viji, C.; Loganathan, D.; Sushma, K. S. N.; Stalin, M.
Elsevier B.V., vol. 230, 2023, ISBN: 18770509 (ISSN), (0).
@proceedings{97,
title = {Investigation of Task Scheduling in Cloud Computing by using Imperialist Competitive and Crow Search Algorithms},
author = {J. Ravi and N. Rajkumar and C. Viji and D. Loganathan and K. S. N. Sushma and M. Stalin},
doi = {10.1016/j.procs.2023.12.044},
isbn = {18770509 (ISSN)},
year = {2023},
date = {2023-01-01},
journal = {Procedia Computer Science},
volume = {230},
pages = {879-889,},
publisher = {Elsevier B.V.},
abstract = {Cloud Storage is a complex method that is a method of processing and data of a cloud built by duplication of thousands of related devices in a complex manner. The main function of the data processing server is to show how many users are being investigated and provide accurate, efficient and efficient information. Important Algorithm Editing Players in the Cloud Defines the virtual machine (VM) required for this purpose. The role of editing the algorithm reduces the effect of the schedule. Naturally affected algorithms have recently been used to quickly comply with the recent traditional algorithms. Given many consumers of many cloud computing services, many researchers may have a serious explanation that many researchers take and discuss the complex themes of NP. Some sites use imperialist algorithms (ICA) and birds. The purpose of the proposed project is to develop intelligent scientific algorithms that focus on the integration of ICA and CSA to obtain data. CSA is concentrated on the corner of food habits. The crow is looking for his friends to get enough food for today's food. This will help the CSA find suitable VMs for these machines and complete the equipment. Cloud Sim is used to calculate CSA output with minmin and ant algorithms. The simulation results show that the CSA is extra powerful than the MinMin and Ant procedures. © 2023 Elsevier B.V.. All rights reserved.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
Cloud Storage is a complex method that is a method of processing and data of a cloud built by duplication of thousands of related devices in a complex manner. The main function of the data processing server is to show how many users are being investigated and provide accurate, efficient and efficient information. Important Algorithm Editing Players in the Cloud Defines the virtual machine (VM) required for this purpose. The role of editing the algorithm reduces the effect of the schedule. Naturally affected algorithms have recently been used to quickly comply with the recent traditional algorithms. Given many consumers of many cloud computing services, many researchers may have a serious explanation that many researchers take and discuss the complex themes of NP. Some sites use imperialist algorithms (ICA) and birds. The purpose of the proposed project is to develop intelligent scientific algorithms that focus on the integration of ICA and CSA to obtain data. CSA is concentrated on the corner of food habits. The crow is looking for his friends to get enough food for today's food. This will help the CSA find suitable VMs for these machines and complete the equipment. Cloud Sim is used to calculate CSA output with minmin and ant algorithms. The simulation results show that the CSA is extra powerful than the MinMin and Ant procedures. © 2023 Elsevier B.V.. All rights reserved.
34.
Rudra, B. B.; Murtugudde, G.
Hybrid Feature Selection with Parallel Multi-Class Support Vector Machine for Land Use Classification Journal Article
In: International Journal of Intelligent Engineering and Systems, vol. 15, pp. 85-94,, 2022, ISBN: 2185310X (ISSN), (1).
@article{110,
title = {Hybrid Feature Selection with Parallel Multi-Class Support Vector Machine for Land Use Classification},
author = {B. B. Rudra and G. Murtugudde},
doi = {10.22266/IJIES2022.0228.09},
isbn = {2185310X (ISSN)},
year = {2022},
date = {2022-01-01},
journal = {International Journal of Intelligent Engineering and Systems},
volume = {15},
pages = {85-94,},
publisher = {Intelligent Network and Systems Society},
abstract = {Land use classification in remote sensing is required in various applications like natural resource management, urban mapping and agriculture etc. Existing methods in the Land use classification which has the limitation of overfitting problem due to the improper feature selection in the method. In this research, the hybrid feature selection methods with Parallel Multi-Class Support Vector Machine (MSVM) is proposed to improve the land use classification performance. The UC Merced and AID datasets were applied to validate the performance of the hybrid feature selection method with the parallel MSVM method. The input images were applied in Histogram Equalization to enhance the image quality which removes the artifacts in the preprocessing stage. The Speeded Up Robust Feature (SURF), Local Ternary Pattern (LTP), Discrete Wavelet Transform (DWT) were applied for feature extraction. The extracted features are applied to hybrid feature selection of Particle Swarm Optimization (PSO) and Grey Wolf Optimization (GWO) method to select the relevant features. The hybrid feature selection method has the advantages of good convergence with higher efficiency in search analysis. The PSO model provided good search exploration to find better solution and GWO method has good convergence of local and global solution. The hybrid method has effective exploration and exploitation for the feature selection. The proposed hybrid features with the MSVM method have 99.15 % accuracy and the existing SVM has 94 % accuracy in land use classification. © 2022,International Journal of Intelligent Engineering and Systems. All Rights Reserved.},
note = {1},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
Land use classification in remote sensing is required in various applications like natural resource management, urban mapping and agriculture etc. Existing methods in the Land use classification which has the limitation of overfitting problem due to the improper feature selection in the method. In this research, the hybrid feature selection methods with Parallel Multi-Class Support Vector Machine (MSVM) is proposed to improve the land use classification performance. The UC Merced and AID datasets were applied to validate the performance of the hybrid feature selection method with the parallel MSVM method. The input images were applied in Histogram Equalization to enhance the image quality which removes the artifacts in the preprocessing stage. The Speeded Up Robust Feature (SURF), Local Ternary Pattern (LTP), Discrete Wavelet Transform (DWT) were applied for feature extraction. The extracted features are applied to hybrid feature selection of Particle Swarm Optimization (PSO) and Grey Wolf Optimization (GWO) method to select the relevant features. The hybrid feature selection method has the advantages of good convergence with higher efficiency in search analysis. The PSO model provided good search exploration to find better solution and GWO method has good convergence of local and global solution. The hybrid method has effective exploration and exploitation for the feature selection. The proposed hybrid features with the MSVM method have 99.15 % accuracy and the existing SVM has 94 % accuracy in land use classification. © 2022,International Journal of Intelligent Engineering and Systems. All Rights Reserved.
35.
Rudra, B. B.; Murtugudde, G.
Remote sensing scene classification using visual geometry group 19 model and multi objective grasshopper optimization algorithm Journal Article
In: International Journal of System Assurance Engineering and Management, vol. 13, pp. 3017-3030,, 2022, ISBN: 09756809 (ISSN), (1).
@article{161,
title = {Remote sensing scene classification using visual geometry group 19 model and multi objective grasshopper optimization algorithm},
author = {B. B. Rudra and G. Murtugudde},
doi = {10.1007/s13198-022-01790-3},
isbn = {09756809 (ISSN)},
year = {2022},
date = {2022-01-01},
journal = {International Journal of System Assurance Engineering and Management},
volume = {13},
pages = {3017-3030,},
publisher = {Springer},
abstract = {Recently, the Remote Sensing Scene Classification (RSSC) has played a vital role in several applications: environment monitoring, urban planning, and land management. The deep neural networks are extensively utilized in the RSSC, because of their superior performance. In recent decades, several scene classification models improve classification accuracy by incorporating extra modules, but it increases the computing overhead and parameters of the models at the inference phase. In addition, the complementarity of the features extracted by the deep learning models is exploited to reduce the improvement of classification accuracy. For addressing the aforementioned issues, a new meta-heuristics based Visual Geometry Group-19 (VGG-19) model is implemented in this research manuscript. After acquiring the aerial images from REmote Sensing Image Scene Classification 45 (RESISC45), Aerial Image Dataset (AID) and the University of California Merced (UC Merced) datasets, the VGG-19 network is applied for classifying the scene categories. In the proposed system, a multi-objective Grasshopper Optimization Algorithm (GOA) is implemented for selecting the optimal hyper-parameters of the VGG-19 model, which helps in reducing the computational complexity and training time of the model. The experimental results demonstrated that the meta-heuristics based VGG-19 model achieved 98.67%, 99.57%, and 98.06% of accuracy on the AID, UC Merced, and RESISC45 datasets, which are superior related to the comparative deep learning models. © 2022, The Author(s) under exclusive licence to The Society for Reliability Engineering, Quality and Operations Management (SREQOM), India and The Division of Operation and Maintenance, Lulea University of Technology, Sweden.},
note = {1},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
Recently, the Remote Sensing Scene Classification (RSSC) has played a vital role in several applications: environment monitoring, urban planning, and land management. The deep neural networks are extensively utilized in the RSSC, because of their superior performance. In recent decades, several scene classification models improve classification accuracy by incorporating extra modules, but it increases the computing overhead and parameters of the models at the inference phase. In addition, the complementarity of the features extracted by the deep learning models is exploited to reduce the improvement of classification accuracy. For addressing the aforementioned issues, a new meta-heuristics based Visual Geometry Group-19 (VGG-19) model is implemented in this research manuscript. After acquiring the aerial images from REmote Sensing Image Scene Classification 45 (RESISC45), Aerial Image Dataset (AID) and the University of California Merced (UC Merced) datasets, the VGG-19 network is applied for classifying the scene categories. In the proposed system, a multi-objective Grasshopper Optimization Algorithm (GOA) is implemented for selecting the optimal hyper-parameters of the VGG-19 model, which helps in reducing the computational complexity and training time of the model. The experimental results demonstrated that the meta-heuristics based VGG-19 model achieved 98.67%, 99.57%, and 98.06% of accuracy on the AID, UC Merced, and RESISC45 datasets, which are superior related to the comparative deep learning models. © 2022, The Author(s) under exclusive licence to The Society for Reliability Engineering, Quality and Operations Management (SREQOM), India and The Division of Operation and Maintenance, Lulea University of Technology, Sweden.
36.
Rajanikanth, P.; Reddy, K. S.
An Efficient Routing Mechanism for Node Localization, Cluster Based Approach and Data Aggregation to Extend WSN Lifetime Journal Article
In: International Journal of Intelligent Engineering and Systems, vol. 15, pp. 305-317,, 2022, ISBN: 2185310X (ISSN), (9).
@article{144,
title = {An Efficient Routing Mechanism for Node Localization, Cluster Based Approach and Data Aggregation to Extend WSN Lifetime},
author = {P. Rajanikanth and K. S. Reddy},
doi = {10.22266/IJIES2022.0228.28},
isbn = {2185310X (ISSN)},
year = {2022},
date = {2022-01-01},
journal = {International Journal of Intelligent Engineering and Systems},
volume = {15},
pages = {305-317,},
publisher = {Intelligent Network and Systems Society},
abstract = {The Last decade the demand for wireless communication has increased massively. Wireless Sensor Networks (WSN) have turned out as a potential solution for various real-time applications namely environmentalmonitoring, health and military applications. These networks follow the random placement strategy hence, identifyingthe location of sensor nodes is a tedious task. Moreover, these networks suffer from various issues such as limitedpower resources. This work, focusses on the sensor node localization and network lifetime enhancement of WirelessSensor Networks (WSN). In localization, security is an important issue, so a trust-based model to secure localizationis investigated in this work. Further, energy aware clustering and Cluster Head (CH) selection has been undertaken toprolong the lifetime of the network. Finally, a data aggregation model which decreases the network energy depletionby discarding the redundant data is presented. The outcome of the model which is proposed has been assessed andexpressed in the form of localization error, energy depletion and alive node. The comparative analysis shows that thelocalization error for varied anchor nodes is obtained as 1.51, 2.28, 2.52 and 1.25 using existing techniques DMA(Distance Mapping Algorithm), MDS-Map (Multidimensional Scaling), DV-Hop (Distance Vector-Hop) andProposed Approach, respectively. Similarly, the average energy consumption is obtained as 1.02, 1.67, 1.90 and 0.82using DMA (Distance Mapping Algorithm), MDS-Map (Multidimensional Scaling), DV-Hop (Distance Vector-Hop)and Proposed Hybrid Approach. The comparative analysis shows that proposed hybrid approach of localization, CH(Cluster Head) formation and aggregation attains better performance when compared with existing techniques inWSN(Wireless Sensor Networks) © 2022,International Journal of Intelligent Engineering and Systems. All Rights Reserved.},
note = {9},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
The Last decade the demand for wireless communication has increased massively. Wireless Sensor Networks (WSN) have turned out as a potential solution for various real-time applications namely environmentalmonitoring, health and military applications. These networks follow the random placement strategy hence, identifyingthe location of sensor nodes is a tedious task. Moreover, these networks suffer from various issues such as limitedpower resources. This work, focusses on the sensor node localization and network lifetime enhancement of WirelessSensor Networks (WSN). In localization, security is an important issue, so a trust-based model to secure localizationis investigated in this work. Further, energy aware clustering and Cluster Head (CH) selection has been undertaken toprolong the lifetime of the network. Finally, a data aggregation model which decreases the network energy depletionby discarding the redundant data is presented. The outcome of the model which is proposed has been assessed andexpressed in the form of localization error, energy depletion and alive node. The comparative analysis shows that thelocalization error for varied anchor nodes is obtained as 1.51, 2.28, 2.52 and 1.25 using existing techniques DMA(Distance Mapping Algorithm), MDS-Map (Multidimensional Scaling), DV-Hop (Distance Vector-Hop) andProposed Approach, respectively. Similarly, the average energy consumption is obtained as 1.02, 1.67, 1.90 and 0.82using DMA (Distance Mapping Algorithm), MDS-Map (Multidimensional Scaling), DV-Hop (Distance Vector-Hop)and Proposed Hybrid Approach. The comparative analysis shows that proposed hybrid approach of localization, CH(Cluster Head) formation and aggregation attains better performance when compared with existing techniques inWSN(Wireless Sensor Networks) © 2022,International Journal of Intelligent Engineering and Systems. All Rights Reserved.
37.
Chavan, P.; Reddy, K. S.
Integrated cross layer optimization approach for quality of service enhancement in wireless network Journal Article
In: Indian Journal of Computer Science and Engineering, vol. 12, pp. 885-898,, 2021, ISBN: 09765166 (ISSN), (5).
@article{158,
title = {Integrated cross layer optimization approach for quality of service enhancement in wireless network},
author = {P. Chavan and K. S. Reddy},
doi = {10.21817/indjcse/2021/v12i4/211204144},
isbn = {09765166 (ISSN)},
year = {2021},
date = {2021-01-01},
journal = {Indian Journal of Computer Science and Engineering},
volume = {12},
pages = {885-898,},
publisher = {Engg Journals Publications},
abstract = {Wireless network has emerged as the primary technology for the next generations networks and it adopts various feature such as ease of deployment, low cost, increased coverage which forms the self-organized structure without relying on any fixed network. Wireless network possesses various multimedia transmission application such as real time delivery of audio, video and VOIP; out of these services video transmission is considered as more challenging as it requires high QoS as the QoS are degraded often due to unreliable nature of network, fading radio signals, node’s mobility. Moreover, high quality video transmission requires high QoS such as efficient bandwidth utilization, high throughput, low delay and high PSNR value for video transmission; in order to utilize the network resources, guaranteed QoS is essential requirement which has been researched by several researchers, however main issue of these approaches are missing end-to-end QoS guarantee for video transmission. In this paper we design and develop an ICLO (Integrated Cross Layer optimization) approach which aims to enhance the Quality of Service; considering two major issue i.e. video encoding and packet transmission. Further, these problem are solved through various optimization model such as optimization of channel modelling, obstacle aware interference modelling, queueing model and video distortion model. Further these optimization are integrated to solve the designed problem which enhances the Quality of Service. Further, we evaluate the ICLO considering the performance metrics like bandwidth efficiency, PSNR, Throughput and End-to-End delay. © 2021, Engg Journals Publications. All rights reserved.},
note = {5},
keywords = {ISE},
pubstate = {published},
tppubtype = {article}
}
Wireless network has emerged as the primary technology for the next generations networks and it adopts various feature such as ease of deployment, low cost, increased coverage which forms the self-organized structure without relying on any fixed network. Wireless network possesses various multimedia transmission application such as real time delivery of audio, video and VOIP; out of these services video transmission is considered as more challenging as it requires high QoS as the QoS are degraded often due to unreliable nature of network, fading radio signals, node’s mobility. Moreover, high quality video transmission requires high QoS such as efficient bandwidth utilization, high throughput, low delay and high PSNR value for video transmission; in order to utilize the network resources, guaranteed QoS is essential requirement which has been researched by several researchers, however main issue of these approaches are missing end-to-end QoS guarantee for video transmission. In this paper we design and develop an ICLO (Integrated Cross Layer optimization) approach which aims to enhance the Quality of Service; considering two major issue i.e. video encoding and packet transmission. Further, these problem are solved through various optimization model such as optimization of channel modelling, obstacle aware interference modelling, queueing model and video distortion model. Further these optimization are integrated to solve the designed problem which enhances the Quality of Service. Further, we evaluate the ICLO considering the performance metrics like bandwidth efficiency, PSNR, Throughput and End-to-End delay. © 2021, Engg Journals Publications. All rights reserved.
38.
Soubhagyalakshmi, P.; Reddy, Satyanarayan
SWOT Analysis of BYOD (Bring Your Own Device) Proceedings
Springer Science and Business Media Deutschland GmbH, vol. 164, 2021, ISBN: 23673370 (ISSN); 978-981159773-2 (ISBN), (1).
@proceedings{111,
title = {SWOT Analysis of BYOD (Bring Your Own Device)},
author = {P. Soubhagyalakshmi and Satyanarayan Reddy},
doi = {10.1007/978-981-15-9774-9_63},
isbn = {23673370 (ISSN); 978-981159773-2 (ISBN)},
year = {2021},
date = {2021-01-01},
journal = {Lecture Notes in Networks and Systems},
volume = {164},
pages = {681-688,},
publisher = {Springer Science and Business Media Deutschland GmbH},
abstract = {Bring your own device (BYOD) grants delegates to use their own gadgets, application, and framework for business or work reason, with an idea of extending productivity, reduces the IT costs, and adjusting undertaking system. A SWOT assessment of BYOD is furthermore a result to provide a succinct audit of favorable circumstances with downsides of BYOD technology. Furthermore, an examination of a modification concept is given which explains a comprehensive diagram of various needs of IT industry and their unmistakable security level over their data and structure for executing BYOD. At the present time, we are focusing on conduction of SWOT examination of BYOD. © 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.},
note = {1},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
Bring your own device (BYOD) grants delegates to use their own gadgets, application, and framework for business or work reason, with an idea of extending productivity, reduces the IT costs, and adjusting undertaking system. A SWOT assessment of BYOD is furthermore a result to provide a succinct audit of favorable circumstances with downsides of BYOD technology. Furthermore, an examination of a modification concept is given which explains a comprehensive diagram of various needs of IT industry and their unmistakable security level over their data and structure for executing BYOD. At the present time, we are focusing on conduction of SWOT examination of BYOD. © 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
39.
Modi, D.; Sutagundar, A. V.; Yalavigi, V.; Aravatagimath, A.
Crop Recommendation Using Machine Learning Algorithm Proceedings
Institute of Electrical and Electronics Engineers Inc., 2021, ISBN: 978-166540341-2 (ISBN), (22).
@proceedings{128,
title = {Crop Recommendation Using Machine Learning Algorithm},
author = {D. Modi and A. V. Sutagundar and V. Yalavigi and A. Aravatagimath},
doi = {10.1109/ISCON52037.2021.9702392},
isbn = {978-166540341-2 (ISBN)},
year = {2021},
date = {2021-01-01},
journal = {2021 5th International Conference on Information Systems and Computer Networks, ISCON 2021},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {Agriculture is extremely important to India's economy and employment. The most common issue faced by Indian farmers is that farmers do not select the appropriate crop for their soil. As a result, productivity is harmed. Agriculture is the main source of income and the backbone of our economy. The poor crop selection has reduced crop production and food shortages across the country which resulted in an increase in farmer suicide. Farmers' problems have been handled by recommendation of suitable crop before sowing. To overcome these issues it is necessary to analyze the soil parameters. This proposed work presents the SVM algorithm based crop recommendation system for the formers. In this work, it is necessary to analyze the profit of the particular crop, which eliminates the loss for the farmers and increase the productivity. SVM algorithm is used for classification to classify the different parameters of the soil and predict the most suitable crop.The proposed algorithm is simulated in anaconda navigator to analyze the soil parameters and recommend a suitable crop. The SVM algorithm is considered for classification. To test the effectiveness of the proposed algorithm accuracy and confusion matrix are computed. © 2021 IEEE.},
note = {22},
keywords = {ISE},
pubstate = {published},
tppubtype = {proceedings}
}
Agriculture is extremely important to India's economy and employment. The most common issue faced by Indian farmers is that farmers do not select the appropriate crop for their soil. As a result, productivity is harmed. Agriculture is the main source of income and the backbone of our economy. The poor crop selection has reduced crop production and food shortages across the country which resulted in an increase in farmer suicide. Farmers' problems have been handled by recommendation of suitable crop before sowing. To overcome these issues it is necessary to analyze the soil parameters. This proposed work presents the SVM algorithm based crop recommendation system for the formers. In this work, it is necessary to analyze the profit of the particular crop, which eliminates the loss for the farmers and increase the productivity. SVM algorithm is used for classification to classify the different parameters of the soil and predict the most suitable crop.The proposed algorithm is simulated in anaconda navigator to analyze the soil parameters and recommend a suitable crop. The SVM algorithm is considered for classification. To test the effectiveness of the proposed algorithm accuracy and confusion matrix are computed. © 2021 IEEE.
40.
D’Souza, S. M.; Reddy, K. S.; Nanda, P.
A survey on guiding customer purchasing patterns in modern consumerism scenario in india Book Chapter
In: vol. 132, pp. 115-126,, Springer, 2021, ISBN: 23673370 (ISSN), (0).
@inbook{130,
title = {A survey on guiding customer purchasing patterns in modern consumerism scenario in india},
author = {S. M. D’Souza and K. S. Reddy and P. Nanda},
doi = {10.1007/978-981-15-5309-7_12},
isbn = {23673370 (ISSN)},
year = {2021},
date = {2021-01-01},
journal = {Lecture Notes in Networks and Systems},
volume = {132},
pages = {115-126,},
publisher = {Springer},
abstract = {This proposed work is used to simplify and ease the shopping experience for a customer through the user-friendly android application. Generally, in the shopping malls user does happy shopping but feels tired and sick for standing in the long queues for the billing. Each product can be billed by reading the barcode on each product; thus, a single person scanning thousands of products of hundreds and more customers is a tedious task. Thus, we have come up with the mobile application which helps the customer do easy shopping without long queues for billing. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021.},
note = {0},
keywords = {ISE},
pubstate = {published},
tppubtype = {inbook}
}
This proposed work is used to simplify and ease the shopping experience for a customer through the user-friendly android application. Generally, in the shopping malls user does happy shopping but feels tired and sick for standing in the long queues for the billing. Each product can be billed by reading the barcode on each product; thus, a single person scanning thousands of products of hundreds and more customers is a tedious task. Thus, we have come up with the mobile application which helps the customer do easy shopping without long queues for billing. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021.
