Ayshwarya Balakumar
Research Subjects
Ayshwarya Balakumar has a total of 44 co-authors. Here are the top co-authors by number of shared publications:
- Irina Nikolayevna Odarich
Togliatti State University (TSU), Tolyatti
2 shared publications - Kathiresan Shankar
Department of Computer Science and Engineering, Saveetha School of Engineering, Saveetha Institute of Medical and Technical Sciences, Chennai, India
2 shared publications S. Senthil
2 shared publicationsThanh Phong
Togliatti state university, Russia
2 shared publications- Wahidah Hashim
College of Computing and Informatics, Universiti Tenaga Nasional (UNITEN), Malaysia
2 shared publications - A. Stephen
Kristu Jayanti (Deemed to be) University
1 shared publication
Latest Publications
- 2 January 2026
- Published by Springer Nature
- p. 485-497
Enhancing Home Security with IoT-Enabled Smart Intruder Detection System on the Cloud
- 2 January 2026
- Published by Springer Nature
- p. 485-497
Enhancing Home Security with IoT-Enabled Smart Intruder Detection System on the Cloud
- Open Access
- 1 October 2025
- Published by Institute of Advanced Engineering and Science in Bulletin of Electrical Engineering and Informatics
- Vol. 14 (5) , 3671-3682
An optimized deep learning framework based on LEE for real time student performance prediction in educational data
Predicting student performance in real-time remains a critical challenge in educational data mining (EDM), especially with large, noisy, and high-dimensional datasets. This study proposes an advanced deep learning framework that integrates learning entropy estimation (LEE) with models such as support vector machines (SVM), you only look once (YOLO), recurrent convolutional neural networks (RCNN), and artificial neural networks (ANN) to enhance feature selection and classification accuracy. The framework follows a systematic pipeline involving data preprocessing, LEE-based feature extraction, and model training on a real-time academic dataset comprising student demographics, attendance, and performance metrics. Among the proposed models, the LEE-based YOLO (LBYOLO) achieved the highest testing accuracy of 93% and the fastest execution time of 1.84 seconds, while the LEE-based ANN (LBANN) demonstrated consistent performance across precision, recall, and F1-score. The results confirm the superiority of deep learning methods over traditional machine learning techniques for educational prediction tasks. This approach enables early detection of at-risk students and supports timely, data-driven educational interventions. Future work will focus on adaptive learning systems and multi-platform student behavior analysis to support personalized education strategies.
- 1 October 2025
- Published by Institute of Advanced Engineering and Science in Bulletin of Electrical Engineering and Informatics
- Vol. 14 (5) , 3671-3682
Open AccessAn optimized deep learning framework based on LEE for real time student performance prediction in educational data
Predicting student performance in real-time remains a critical challenge in educational data mining (EDM), especially with large, noisy, and high-dimensional datasets. This study proposes an advanced deep learning framework that integrates learning entropy estimation (LEE) with models such as support vector machines (SVM), you only look once (YOLO), recurrent convolutional neural networks (RCNN), and artificial neural networks (ANN) to enhance feature selection and classification accuracy. The framework follows a systematic pipeline involving data preprocessing, LEE-based feature extraction, and model training on a real-time academic dataset comprising student demographics, attendance, and performance metrics. Among the proposed models, the LEE-based YOLO (LBYOLO) achieved the highest testing accuracy of 93% and the fastest execution time of 1.84 seconds, while the LEE-based ANN (LBANN) demonstrated consistent performance across precision, recall, and F1-score. The results confirm the superiority of deep learning methods over traditional machine learning techniques for educational prediction tasks. This approach enables early detection of at-risk students and supports timely, data-driven educational interventions. Future work will focus on adaptive learning systems and multi-platform student behavior analysis to support personalized education strategies.
- 27 February 2024
- Published by Springer Nature
- p. 141-154
Heart Failure Prediction for a Patient Using Hybrid African Buffalo Optimization with Naive Bayes Machine Learning Techniques
- 27 February 2024
- Published by Springer Nature
- p. 141-154
Heart Failure Prediction for a Patient Using Hybrid African Buffalo Optimization with Naive Bayes Machine Learning Techniques
- 1 January 2021
- Published by IGI Global Scientific Publishing
- p. 270-287
Edge Computing and IoT Technologies for Medical Applications
Dynamic observation of blood sugar levels is essential for patients diagnosed with diabetes mellitus in order to control the glycaemia. Inevitably, they must accomplish a capillary test three times per day and laboratory test once or twice per month. These regular methods make patients uncomfortable because patients have to prick their finger every time in order to measure the glucose concentration. Modern health monitoring systems rely on IoT. However, the number of advanced IoT-based continuous glucose monitoring systems is small and has several limitations. Here the authors study feasibility of invasive and continuous glucose monitoring system utilizing IoT-based approach. They designed an IoT-based system architecture from a sensor device to a back-end system for presenting real-time data in various forms to end-users. The results show that the system is able to achieve continuous glucose monitoring remotely in real time, and a high level of energy efficiency can be achieved by applying the nRF compound, power management, and energy harvesting unit altogether in the sensor units.
- 1 January 2021
- Published by IGI Global Scientific Publishing
- p. 270-287
Edge Computing and IoT Technologies for Medical Applications
Dynamic observation of blood sugar levels is essential for patients diagnosed with diabetes mellitus in order to control the glycaemia. Inevitably, they must accomplish a capillary test three times per day and laboratory test once or twice per month. These regular methods make patients uncomfortable because patients have to prick their finger every time in order to measure the glucose concentration. Modern health monitoring systems rely on IoT. However, the number of advanced IoT-based continuous glucose monitoring systems is small and has several limitations. Here the authors study feasibility of invasive and continuous glucose monitoring system utilizing IoT-based approach. They designed an IoT-based system architecture from a sensor device to a back-end system for presenting real-time data in various forms to end-users. The results show that the system is able to achieve continuous glucose monitoring remotely in real time, and a high level of energy efficiency can be achieved by applying the nRF compound, power management, and energy harvesting unit altogether in the sensor units.
- 1 January 2021
- Published by IGI Global Scientific Publishing
- p. 55-72
Internet of Things (IoT)
Naturally, IoT network consists of constrained devices with limited power, storage, and processing capabilities. However, IoT has some challenges to face internal and external issues, for instance, security issues, connectivity complex, data management, etc. In such a case, IoT is not supported by a heavyweight protocol suite. So, developing the lightweight protocol suite is the challenge for IoT environment. This chapter also describes such protocols that fulfill the requirements of IoT. In fact, IoT environment consists of a huge amount of devices. The controlling of all the devices is another issue, as well as data analytics among different devices, also considered a major issue. As a constraint, this chapter focuses to design a standard protocol suite for IoT environment.
- 1 January 2021
- Published by IGI Global Scientific Publishing
- p. 55-72
Internet of Things (IoT)
Naturally, IoT network consists of constrained devices with limited power, storage, and processing capabilities. However, IoT has some challenges to face internal and external issues, for instance, security issues, connectivity complex, data management, etc. In such a case, IoT is not supported by a heavyweight protocol suite. So, developing the lightweight protocol suite is the challenge for IoT environment. This chapter also describes such protocols that fulfill the requirements of IoT. In fact, IoT environment consists of a huge amount of devices. The controlling of all the devices is another issue, as well as data analytics among different devices, also considered a major issue. As a constraint, this chapter focuses to design a standard protocol suite for IoT environment.
- 1 January 2020
- Published by IGI Global Scientific Publishing
- p. 176-196
Machine Learning Is the Future for Lung Cancer Prognosis and Prediction
Lung cancer is one of the major reasons for the death if it is not diagnosed in the early stages of cancer. It is the one among the most dreadful disease which affects in the lungs function. It can be identified only after the disease spread into the deeper parts of the lungs and then only it will make a life threading problem. Lung cancer prognosis which was done based on the various parameters such as age, sex, condition of smoking, duration of smoking and count of smoking per day. The proceedings were done using the algorithm for the time to first cigarette after awakening which is represented as TTFC. The expert doctor says that the back-propagation network is a great deal in the recognition of the lung cancer without any involvement by them. This research is based on the classification of lung cancer and its stages using the establishment of the BPN and predicts the recurrence. Similarly, with this BPN, an algorithm that is inspired from its habitat known as ant lion optimization algorithm is also used in the optimization of weights and parameters of the BPN. The use of the ALO algorithm provides an improved convergence mechanism by improving the proposed technique's accuracy. The use of this proposed method with the BPN optimizes the network and the ALO optimizer provides an accurate prediction of the lung cancer by the earlier stage and even predicts the changes for reoccurrence after diagnosis. The prognosis analysis was made by the various comparative study between the characteristic features of HIV and the unaffected person using the algorithm such as the Wilcoxon rank-sum test. This algorithm will continuously classify the viral load and CD4 count which is based on factors such as age, sex, and smoking activities. It will be useful for early diagnosis and future prediction. Lung cancer rates can be analyzed based on the incident rates of affected and unaffected persons to HIV infections.
- 1 January 2020
- Published by IGI Global Scientific Publishing
- p. 176-196
Machine Learning Is the Future for Lung Cancer Prognosis and Prediction
Lung cancer is one of the major reasons for the death if it is not diagnosed in the early stages of cancer. It is the one among the most dreadful disease which affects in the lungs function. It can be identified only after the disease spread into the deeper parts of the lungs and then only it will make a life threading problem. Lung cancer prognosis which was done based on the various parameters such as age, sex, condition of smoking, duration of smoking and count of smoking per day. The proceedings were done using the algorithm for the time to first cigarette after awakening which is represented as TTFC. The expert doctor says that the back-propagation network is a great deal in the recognition of the lung cancer without any involvement by them. This research is based on the classification of lung cancer and its stages using the establishment of the BPN and predicts the recurrence. Similarly, with this BPN, an algorithm that is inspired from its habitat known as ant lion optimization algorithm is also used in the optimization of weights and parameters of the BPN. The use of the ALO algorithm provides an improved convergence mechanism by improving the proposed technique's accuracy. The use of this proposed method with the BPN optimizes the network and the ALO optimizer provides an accurate prediction of the lung cancer by the earlier stage and even predicts the changes for reoccurrence after diagnosis. The prognosis analysis was made by the various comparative study between the characteristic features of HIV and the unaffected person using the algorithm such as the Wilcoxon rank-sum test. This algorithm will continuously classify the viral load and CD4 count which is based on factors such as age, sex, and smoking activities. It will be useful for early diagnosis and future prediction. Lung cancer rates can be analyzed based on the incident rates of affected and unaffected persons to HIV infections.
- Open Access
- 10 August 2019
- Published by Blue Eyes Intelligence Engineering and Sciences Engineering and Sciences Publication (BEIESP) in International Journal of Recent Technology and Engineering (IJRTE)
- Vol. 8 (2S3) , 1562-1567
Into the Furniture Woods: Analytical Hierarchy Process Method
Sidojaya Furniture is one of the Small and Medium Enterprises that is still developing in improving its business. Related with the lack of information obtained by furniture craftsmen in determining qualified wood, Sidojaya Furniture has problems in determining qualified wood that matches with specified criteria for making furniture quickly and accurately. The objective to be achieved from this study was to create Decision Support System for selecting qualified wood in Sidojaya Furniture using AHP (Analytical Hierarchy Process) Method and using Microsoft Visual Basic programming language that can help provide information about qualified wood for furniture production, so that craftsmen can make decisions quickly to make choice. AHP method is a method of breaking complex/ complicated problem in unstructured situation into component parts. Decision support system for the selection of qualified wood can make it easier for craftsmen to decide on the type and nature of qualified wood and obtain information about the texture, color, direction of fiber, age, durability and growing place of wood to be used. From three alternatives tested, it can be obtained, the decision of qualified wood can be used to make furniture was teak wood with a weight score of 1.389.
- 10 August 2019
- Published by Blue Eyes Intelligence Engineering and Sciences Engineering and Sciences Publication (BEIESP) in International Journal of Recent Technology and Engineering (IJRTE)
- Vol. 8 (2S3) , 1562-1567
Open AccessInto the Furniture Woods: Analytical Hierarchy Process Method
Sidojaya Furniture is one of the Small and Medium Enterprises that is still developing in improving its business. Related with the lack of information obtained by furniture craftsmen in determining qualified wood, Sidojaya Furniture has problems in determining qualified wood that matches with specified criteria for making furniture quickly and accurately. The objective to be achieved from this study was to create Decision Support System for selecting qualified wood in Sidojaya Furniture using AHP (Analytical Hierarchy Process) Method and using Microsoft Visual Basic programming language that can help provide information about qualified wood for furniture production, so that craftsmen can make decisions quickly to make choice. AHP method is a method of breaking complex/ complicated problem in unstructured situation into component parts. Decision support system for the selection of qualified wood can make it easier for craftsmen to decide on the type and nature of qualified wood and obtain information about the texture, color, direction of fiber, age, durability and growing place of wood to be used. From three alternatives tested, it can be obtained, the decision of qualified wood can be used to make furniture was teak wood with a weight score of 1.389.
- Open Access
- 10 August 2019
- Published by Blue Eyes Intelligence Engineering and Sciences Engineering and Sciences Publication (BEIESP) in International Journal of Recent Technology and Engineering (IJRTE)
- Vol. 8 (2S3) , 1507-1512
Design Analysis of Web-Based Employee Information System in Office of Technical Implementation Unit of Education Department
KUPT Education Agency of Gunung Alip Sub-District is institution that handles education of elementary/kindergarten level in Gunung Alip Sub-District and all employees in KUPT Education Agency. KUPT Education Agency of Gunung Alip Sub-District in maintaining administration manually. Therefore Information System is needed to can help manage administration, especially personnel, quickly and easily. In making this information system the author used the waterfall method, Website programming. The resulting information system is expected to be used to manage employee data at the KUPT Education Office in Gunung Alip Sub-District.
- 10 August 2019
- Published by Blue Eyes Intelligence Engineering and Sciences Engineering and Sciences Publication (BEIESP) in International Journal of Recent Technology and Engineering (IJRTE)
- Vol. 8 (2S3) , 1507-1512
Open AccessDesign Analysis of Web-Based Employee Information System in Office of Technical Implementation Unit of Education Department
KUPT Education Agency of Gunung Alip Sub-District is institution that handles education of elementary/kindergarten level in Gunung Alip Sub-District and all employees in KUPT Education Agency. KUPT Education Agency of Gunung Alip Sub-District in maintaining administration manually. Therefore Information System is needed to can help manage administration, especially personnel, quickly and easily. In making this information system the author used the waterfall method, Website programming. The resulting information system is expected to be used to manage employee data at the KUPT Education Office in Gunung Alip Sub-District.
- 30 July 2019
- Published by Blue Eyes Intelligence Engineering and Sciences Engineering and Sciences Publication (BEIESP) in International Journal of Recent Technology and Engineering (IJRTE)
- Vol. 8 (2) , 6368-6391
Improving the Performance of Lung Cancer Detection at Earlier Stage and Prediction of Reoccurrence using the Neural Networks and Ant Lion Optimizer
Lung cancer is considered to be the one among the most dreaded disease which will be the main reason for the death of individuals and having greater deterioration of death if it is not identified at primitive stage. Because of the fact that Lung cancer could be identified only after spreading to the parts of lungs to a greater extent and it is very tough to predict the presence of lung cancer at the earlier stage. Moreover, it involves greater error in the diagnosing the presence of Lung cancer by Radiologists and Expert Doctors. Therefor it is compulsory to design an intelligent and automated system for accurately predicting the cancer and stage at which the stage of cancer or enhancing the accuracy of prediction for detecting the cancer at earlier which will be much helpful in deciding the treatment type and depth of the treatment based on the extent of disease. Currently application of ANN strategies are the influential ways in supporting expert doctor for examining, complicated medical increase across a wider category of medical application. Back Propagation Network are ideal in recognizing lung cancer and there is no requirement involvement by expert doctors. Maximum number of applications of BPN in medical diagnosis will be utilized in the applications related to decision making of the presence or absence of disease; by which the performance will be reliant over the considered features and allocating the patient with minimum number of classes. Here this research paper establishes the idea of using BPN in the classification of the lung cancer and its stages and the predicting the possibility of recurrence. Along with the BPN, a nature inspired Meta Heuristics that is termed as Ant Lion Optimization Algorithm is used in optimizing the parameters and weights of Back Propagation Network. By using the Ant Lion Optimization Algorithm, the convergence mechanism is improved along with improving the accuracy of the proposed technique and it avoids the chance of getting caught within the clutches of local minima. By using this proposed method BPN network optimized with the help of antlion optimizer more accurate prediction of lung cancer is possible even at primitive stage and the predicting of chance of reoccurrence even after undergoing the appropriate treatment.
- 30 July 2019
- Published by Blue Eyes Intelligence Engineering and Sciences Engineering and Sciences Publication (BEIESP) in International Journal of Recent Technology and Engineering (IJRTE)
- Vol. 8 (2) , 6368-6391
Improving the Performance of Lung Cancer Detection at Earlier Stage and Prediction of Reoccurrence using the Neural Networks and Ant Lion Optimizer
Lung cancer is considered to be the one among the most dreaded disease which will be the main reason for the death of individuals and having greater deterioration of death if it is not identified at primitive stage. Because of the fact that Lung cancer could be identified only after spreading to the parts of lungs to a greater extent and it is very tough to predict the presence of lung cancer at the earlier stage. Moreover, it involves greater error in the diagnosing the presence of Lung cancer by Radiologists and Expert Doctors. Therefor it is compulsory to design an intelligent and automated system for accurately predicting the cancer and stage at which the stage of cancer or enhancing the accuracy of prediction for detecting the cancer at earlier which will be much helpful in deciding the treatment type and depth of the treatment based on the extent of disease. Currently application of ANN strategies are the influential ways in supporting expert doctor for examining, complicated medical increase across a wider category of medical application. Back Propagation Network are ideal in recognizing lung cancer and there is no requirement involvement by expert doctors. Maximum number of applications of BPN in medical diagnosis will be utilized in the applications related to decision making of the presence or absence of disease; by which the performance will be reliant over the considered features and allocating the patient with minimum number of classes. Here this research paper establishes the idea of using BPN in the classification of the lung cancer and its stages and the predicting the possibility of recurrence. Along with the BPN, a nature inspired Meta Heuristics that is termed as Ant Lion Optimization Algorithm is used in optimizing the parameters and weights of Back Propagation Network. By using the Ant Lion Optimization Algorithm, the convergence mechanism is improved along with improving the accuracy of the proposed technique and it avoids the chance of getting caught within the clutches of local minima. By using this proposed method BPN network optimized with the help of antlion optimizer more accurate prediction of lung cancer is possible even at primitive stage and the predicting of chance of reoccurrence even after undergoing the appropriate treatment.
- Retracted
- 1 August 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 210-216
Retracted: Predicting Lung Cancer Using Datamining Techniques With the AID of SVM Classifier
The suggested techniques provide a noble quality tool to predict lung tumor classification and play a major role, particularly in the finding and classification of medical data. The literature reports a number of lung cancer diagnosis systems which predict normal and abnormal lung cancers with the support of SVM. Our proposed research focuses on predicting lung cancer whether it is normal or abnormal, with respect to the classification technique. Initially, in the preprocessing phase, suitable data from the input data set are extracted after preprocessing; the resultant output is fed to the feature selection. In this feature selection phase, the features are selected with the aid of the firefly algorithm. After the feature selection, the particular features are served in to the support vector machine (SVM) classifier; with the aid of this classifier, the data are classified as either normal or abnormal. The proposed method will be implemented in Matlab with various lung cancer data. In addition to this, our proposed work will be in comparison with the present strategies and algorithms for proving that our proposed work is the best one.
- 1 August 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 210-216
RetractedRetracted: Predicting Lung Cancer Using Datamining Techniques With the AID of SVM Classifier
The suggested techniques provide a noble quality tool to predict lung tumor classification and play a major role, particularly in the finding and classification of medical data. The literature reports a number of lung cancer diagnosis systems which predict normal and abnormal lung cancers with the support of SVM. Our proposed research focuses on predicting lung cancer whether it is normal or abnormal, with respect to the classification technique. Initially, in the preprocessing phase, suitable data from the input data set are extracted after preprocessing; the resultant output is fed to the feature selection. In this feature selection phase, the features are selected with the aid of the firefly algorithm. After the feature selection, the particular features are served in to the support vector machine (SVM) classifier; with the aid of this classifier, the data are classified as either normal or abnormal. The proposed method will be implemented in Matlab with various lung cancer data. In addition to this, our proposed work will be in comparison with the present strategies and algorithms for proving that our proposed work is the best one.