Cataloging of Alzheimer’s Disease and Its Stages Using Machine Learning S. Rajalakshmi, L. Malliga, P. Mathiyalagan, P. Prema, S. Biruntha, K. Rama Abirami Decision Sciences in Bioinformatics Theory and Practice, 2026 Alzheimer’s disease is a form of dementia and is classified as a cognitive disorder. In this we focus on this condition and explore various approaches to manage it. Feature extraction is the major challenging factor in managing and forecasting a huge number of datasets, but the major problem faced here is that it is unable to organize and concentrate on the precise highlights from the datasets. The objective is to develop a method that looks for amyloid-based biomarkers, which can be used for identifying early signs of AD using artificial intelligence. Furthermore, the proposed approach has demonstrated exceptional performance compared to traditional AI techniques in recognizing complex and ambiguous patterns within multi-layered data. The use of AI in the early diagnosis and the automated management of Alzheimer’s disease (AD) has recently attracted significant attention.
Vox Civitas: AI-Powered Civic Action S. Biruntha, Nanditha Noble, A.L. Sudarrshana, R. Sureshkumar 2025 IEEE International Conference on Communication Networks and Computing Cnc 2025, 2025 Although artificial intelligence has advanced rapidly, there are still low participation levels, biases, and inability to scale up with civic participation and policy discussion, most digital participation platforms have less than 35 percent response rates, and fewer incorporate citizen views into practical policy results. This paper is a research proposal of an AI-based civic action system, incorporating large language models (LLMs), natural language processing (NLP), and computational social choice algorithms to enable civic deliberation that is large and inclusive, and data-driven. The framework was experimented with 3,000 simulated citizens profile based on the knowledge of participatory design, AI governance, and algorithmic fairness by applying sentiment analysis, topic clustering, and evaluation of the quality of the argument to mediate online discussion. In all stages of simulation and alignment, we used the GPT-3.5-Turbo model as our base LLM. The 3,000 synthetic citizen profiles were generated by conditioning GPT-3.5-Turbo on demographic, policy-interest, and sentiment constraints, and cross-validated with accuracy of 90% using a publicly available participatory-budgeting dataset (OpenPB-2022) to guarantee plausible distribution." The experimental findings show that the accuracy of consensus detection was improved by 27%, the moderate time was diminished by 41%, and the F1-score was 0.89 to detect a common-ground statement, as well as 22% of fairness and inclusivity when using the tool compared to the standard deliberation tools. Our improvements over a RoBERTa-large baseline for moderation and a TF-IDF consensus voter were statistically significant (paired t-test, p<0.05) across 5 runs. These results affirm that it is true that AI-human facilitation can be effectively scaled up to deliberative democracy, without impairing the level of representation or transparency. Expanding on it, the paper presents a new concept, the CivicGPT, the dynamic deliberation engine, that combines the work of the LLM and participatory budgeting with fairness optimization, which is a radical change in the system of ethical and transparent and participatory decisions made by the population.
Machine Learning System for Predicting Latency in Next-Generation Wireless Network Vivekanandhan V, K.Shanthi, Nandhini N, S.Biruntha, K Kiran, P. Chozha Rajan 2025 IEEE 3rd Global Conference on Wireless Computing and Networking Gcwcn 2025, 2025 The upcoming wireless networks 5G and 6G need extremely short delays to support essential applications such as autonomous systems together with real-time voice and data communications. The precise prediction of latency needs to be established for wireless networks to optimize resources and improve user satisfaction. The research develops a machine learning prediction system for wireless networks' latency in heterogeneous environments through Random Forest and XGBoost and Long Short-Term Memory (LSTM) network implementations. The predictive system integrates signal-to- noise ratio (SNR) along with bandwidth into its analysis besides using user mobility parameters and base station loading metrics and channel environmental data. The LSTM model demonstrated excellent results during testing of simulated and genuine datasets along with reaching an R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> score of 0.93 and a Mean Absolute Error (MAE) of 1.83 ms above traditional regression models. The research findings show that the model successfully predicts latency measurements using precise accuracy across different network operational conditions. This anticipatory system creates a flexible real-time scheduling solution for 5G and 6G networks which controls ultra-reliable low-latency communication (URLLC) applications efficiently.
Federated Optimization Algorithms S. Biruntha, S. Rajalakshimi, M. Kavitha Federated Learning Unlocking the Power of Collaborative Intelligence, 2024 Federated learning (FL), a distributed machine-learning framework, has been widely used in a number of industries recently and is positioned to successfully preserve data security and privacy. However, the system's heterogeneity and the statistical heterogeneity of FL provide significant challenges to the quality of the global model. In the context of FL system resource efficiency, this chapter looks into server and client resource allocation and provides an optimization method. This method utilizes all of the server's processing power to determine the ideal weight value for every client by fusing federated learning with adaptive learning. This method greatly reduces the negative impacts of statistical and system heterogeneity by aggregating the global model based on the ideal weight value. In FL, micro-batch gradient descent is typically used to optimize model weights, and it seems to be particularly effective in federated settings. Numerous adjustments have been suggested for conventional machine learning setups in an attempt to speed up the learning process and assist overcome the difficulties caused by the large dimensionality and nonconvexity of the parameter search spaces.
Cyber-Physical Energy Systems for Electric Vehicles ShaikMahaboob Basha, P. Akilandeswari, M. Suguna, D. Prakash, S. Biruntha, P. Vivekanandan Cyber Physical Energy Systems, 2024 This study presents a cyber-physical power system framework for an electric vehicle prototype (EVP). It utilizes a genetic algorithm (GA-NSGAII) to maximize the use of multiple power sources, including kinetic energy, potential energy, and chemical energy. The proposed method can handle diverse power sources and adapt to various conditions while detecting potential anomalies in the electro-mechanical transfer process. The study also introduces a multiple-switch converter model that enables flexible power source connections and efficient terminal voltage control in a multi-source network. The approach demonstrates reduced energy consumption, improved vehicle performance, and enhanced construction. A conceptual proof is provided using VIRTUOSE urban electric vehicle models. The main barrier to the widespread adoption of electric cars is the high cost of storage units, such as fuel cells. This study offers a straightforward strategy to overcome this financial obstacle that is more efficient and easier to incorporate than existing research methods. In conclusion, this research proposes a cyber-physical power management structure that integrates powerful algorithms to handle diverse power sources for innovative electric vehicles. It includes anomaly detection, simulation of multiple-switching converters, and effective voltage regulation in multi-source systems.
Automatic Detection of Diabetic Retinopathy on Digital Fundus Image S. Biruntha, R.P. Narmadha 8th International Conference on Advanced Computing and Communication Systems Icaccs 2022, 2022 Eye disease is a prevalent health problem that affects people all over the world. Diabetic Retinopathy (DR) and Glaucoma are two such diseases. Diabetic Retinopathy (DR) is the damage to the retina of the eye caused by diabetes complications. Diabetic Retinopathy affects a huge number of people and can cause blurring of vision or perhaps blindness if not treated early. As a result, it is critical to recognise DR early and treat it, as failing to do so may result in vision blurring or total blindness. Glaucoma is an eye disease in which the optic nerve, which transmits information from the eye to the brain, is damaged. Glaucoma has no medical symptoms in its early stages, which is why it is so hazardous; by the time you discover vision difficulties, the disease has progressed to irreversible blindness. It can be fatal if left untreated or unchecked.
Rainfall Prediction using kNN and Decision Tree S Biruntha, B S Sowmiya, R Subashri, M Vasanth Proceedings of the International Conference on Electronics and Renewable Systems Icears 2022, 2022 Rainfall forecasting is extremely important in a variety of situations and contexts. By implementing good security precautions in advance, it is possible to significantly limit the consequences of unexpected and excessive rains. Accurate rainfall forecasts have become more difficult than ever before due to climatic changes. Data mining algorithms can forecast rainfall by identifying hidden patterns in meteorological variables from previous data. This study contributes by investigating the application of two data mining approaches for rainfall prediction in the city of Austin. k Nearest Neighbour (kNN) and Decision Trees are some of the techniques used. The dataset comes from a weather forecasting service and includes numerous atmospheric parameters. The pre-processing approach, which includes cleaning and normalising operations, is utilised for successful prediction. The performance of data mining algorithms are evaluated in terms of accuracy, recall, and f-measure with varied training/test data ratios. The future year's rainfall is estimated using the Decision Tree and kNN machine learning algorithms and compare the results obtained by each approach.
Retraction Notice: Distinguishing Reviews Through Sentiment Analysis Using Machine Learning Techniques (2022 8th International Conference on Advanced Computing and Communication Systems (ICACCS) DOI: 10.1109/ICACCS54159.2022.9785354) S Biruntha, Arul Ganeshan S, Ashwin B, Padmasankar K S 8th International Conference on Advanced Computing and Communication Systems Icaccs 2022, 2022 Opinion Mining is a technique of automated extraction of information from the opinion of people about any certain subject or problem. The feasibility of Opinion mining and Sentiment Analysis tool is to “process a bunch of indexed lists for a given thing, creating a rundown of item credits (quality, highlights and so forth) and accumulating opinion”. Yet, with the progression of time additional fascinating applications and improvements appeared around here and presently its fundamental objective is to make PC ready to perceive and produce feelings like human. This paper will attempt to zero in on the essential meanings of Opinion Mining, investigation and Sentiment Analysis is the name given to this new field of study. In recent years, scientists have come up with a number of solutions to the problem. Information Retrieval (IR) and Natural Language Processing (NLP) interact in the subject of Opinion Mining and Sentiment Analysis, which has a few distinct trains, such as message mining and Information Extraction. Sentiment Analysis may be performed using a variety of NLP tools, all of which this paper aimed to demonstrate, from basic definitions to a broad range of applications. Recently, it's been a really active exploring zone. In fact, it has made its way into everything from board science to software engineering of etymological assets needed for Opinion Mining, scarcely any AI methods based on their utilization and significance for the examination, assessment of Sentiment arrangements and its different applications. They are attempting to get assessment data to investigate and sum up the suppositions communicated naturally with PCs. Opinion Mining
IoT based border alerting system for fishermen Journal of Advanced Research in Dynamical and Control Systems, 2018
Techniques on text mining M. Sukanya, S. Biruntha Proceedings of 2012 IEEE International Conference on Advanced Communication Control and Computing Technologies Icaccct 2012, 2012
