Cyber Hacking Breaches Prediction Using Machine Learning
Authors: Dr. Prabakaran Narayan, B.Vamshi, M.Supraja, B.Rama Sudha, K. Vijay Sai
Abstract: Cyber-physical systems (cps) have made significant progress in many dynamic ap-plications due to the integration between physical processes, computational resources, and com-munication capabilities. However, cyber-attacks are a major threat to these systems. Unlike faults that occurs by accidents cyber-physical systems, cyber-attacks occur intelligently and stealthy. Some of these attacks which are called deception attacks, inject false data from sensors or controllers, and also by compromising with some cyber components, corrupt data, or enter misinformation into the system. If the system is unaware of the existence of these attacks, it won’t be able to detect them, and performance may be disrupted or disabled altogether. There-fore, it is necessary to adapt algorithms to identify these types of attacks in these systems. It should be noted that the data generated in these systems is produced in very large number, with so much variety, and high speed, so it is important to use machine learning algorithms to facili-tate the analysis and evaluation of data and to identify hidden patterns. In this research, the CPS is model as a network of agents that move in union with each other, and one agent is considered as a leader, and the other agents are commanded by the leader. The proposed method in this study is to use the structure of deep neural networks for the detection phase, which should in-form the systemof the existence of the attack in the initial moments of the attack. The use of re-silient control algorithms in the network to isolate the misbehave agent in the leader-follower mechanism has been investigated. In the presented control method, after the attack detection phase with the use of a deep neural network, the control system uses the reputation algorithm to isolate the misbehave agent. Experimental analysis shows us that deep learning algorithms can detect attacks with higher performance that usual methods and can make cyber security simpler, more proactive, less expensive and far more effective.
Deepfake Detection On Social Media: Leveraging Deep Learning And Fasttext Embeddings For Identifying Machine-Generated Tweets
Authors: Dr. M. V. Subba Reddy, Nagularpu Hepsiba, Vaddemani Yagna Sree, Patnam Chandra Moulieswar, Shaik Asif
Abstract: Opinions on social media can be swayed thanks to new developments in natural language crea-tion. The capacity of deep neural networks to generate content has also been enhanced via lan-guage modeling. Because of this, text-generative algorithms have improved to the point that at-tackers may train social bots to publish deepfakes that seem legitimate and sway public opinion. Reliable and precise deepfake social media message detection systems are required to address this issue. Keeping this in mind, the present research finds Twitter posts that are made by machines. Using a basic deep learning model and word embeddings, this study leverages the publicly avail-able Tweepfake dataset to distinguish between human and bot-generated tweets. A standard con-volutional neural network (CNN) architecture is trained to detect deepfake tweets using FastText word embeddings. In order to prove that the suggested strategy is better, this research compared it to various machine learning models that served as baselines. Here are some baseline approaches: FastText, Term Frequency, FastText subword embeddings, and Term Frequency Inverse Docu-ment Frequency. We also compare the proposed method to other deep learning models, such as CNN-LSTM and Long short-term memory (LSTM), to demonstrate its efficacy and utility in solving the problem. The CNN architecture, when combined with FastText embeddings, efficient-ly and correctly classifies twitter data with a 93% accuracy rate, according to the experimental results.
AI-Powered Smart Crop Advisory and Monitoring Platform
Authors: Mrs.K.M.Swarna Devi, Prabavathi V, Santhiya S, Thamizharasi S
Abstract: Agriculture faces increasing challenges due to climate variability, resource limitations, and the need for sustainable productivity. This paper presents an AI-Powered Smart Crop Advisory and Monitoring Platform that leverages artificial intelligence and data analytics to support informed agricultural decision-making. The system analyzes historical crop data, soil characteristics, weath-er patterns, and satellite imagery to assess crop health and growth conditions. Machine learning models generate accurate recommendations for irrigation planning, fertilizer management, pest and disease identification, and yield prediction. Image processing and computer vision techniques enable early detection of crop stress and diseases, reducing potential losses. The platform pro-vides timely, location-specific advisory services to farmers, improving crop quality and resource efficiency. By minimizing dependency on manual expertise and enhancing precision farming practices, the proposed solution contributes to increased agricultural productivity, economic sus-tainability, and food security. The system demonstrates the potential of artificial intelligence as a reliable tool for modern, data-driven agriculture.
Secure Data Wiping For Trustworthy IT Asset Recycling
Authors: Kanimozhi S, Priyadharshini N, Priyanka D, Sujitha M
Abstract: Secure data wiping plays a vital role in enabling trustworthy IT asset recycling by ensuring that confidential and sensitive data is permanently erased from electronic devices at the end of their lifecycle. As organizations increasingly upgrade and dispose of IT infrastructure, improper data removal poses serious risks, including data leakage, privacy violations, and non-compliance with data protection regulations. This work proposes a robust secure data wiping framework designed to address these challenges through standardized and verifiable data erasure techniques. The framework supports multiple storage technologies, including hard disk drives and solid-state drives, and applies recognized international data sanitization standards to guarantee complete data destruction. Automated verification mechanisms and detailed audit trails are incorporated to pro-vide transparency, traceability, and compliance assurance throughout the recycling process. In addition, the framework integrates secure data wiping with responsible IT asset recycling practic-es, ensuring that devices can be safely reused, refurbished, or recycled without compromising data security. The proposed approach enhances organizational trust, reduces environmental im-pact, and supports sustainable e-waste management while maintaining strict data protection and regulatory compliance.
A Multi-Perspective Fraud Detection Method for multiparticipante-Commercetransactions
Authors: Limbakar Manjula bai, D. Indhu, D. Sravani, G. Suvarna, K.Kishore Reddy, B.MohanReddy
Abstract: In the realm of e-commerce, where transactions involve multiple participants such as buyers, sellers, and intermediaries, the detection of fraudulent activities presents a significant challenge. To address this issue, our proposed method focuses on a Mult perspective approach aimed at enhancing fraud detection accuracy and efficiency. The first step involves the detection of user behaviors, wherein we leverage various techniques such as behavioral analysis and examination of transaction histories to gain insights into normal user behavior patterns. By understanding typical user interactions within the ecommerce ecosystem, we establish a baseline against which abnormal behaviors can be identified. Subsequently, we investigate into the analysis of abnor-malities for feature extraction. Utilizing sophisticated anomaly detection algorithms, we scrutinize transaction data to uncover irregular patterns indicative of potentially fraudulent activities. This process allows us to extract important features that serve as key indicators for fraud detection. Finally, we employ an ensemble classification model to implement our fraud detection mecha-nism.
Ai-Powered Mentorship And Internship Support Platform
Authors: Dr.S. Manikandan, Chandru S, Dhanush V, Girinath A M
Abstract: AI-Powered Mentorship and Internship Support Platform is an advanced career development framework designed to connect students, mentors, and recruiters through intelligent automation. The system integrates machine learning, natural language processing, and data-driven recommen-dation techniques to deliver personalized mentorship and internship opportunities. It evaluates user profiles, academic records, resumes, and career interests to construct dynamic skill profiles and detect competency gaps. Based on this analysis, the platform generates customized learning pathways, mentor suggestions, and relevant internship matches aligned with individual goals.A similarity-based matching algorithm ensures accurate mentor–mentee pairing, while predictive analytics estimate internship suitability and career readiness. The platform also incorporates an AI-driven chatbot that provides continuous assistance in resume development, interview prepara-tion, and career decision-making. Performance monitoring dashboards track progress, mentor feedback, and internship outcomes, offering actionable insights for improvement. By streamlining mentorship allocation and internship recommendations, the proposed system enhances accessibil-ity, efficiency, and transparency in professional development services. Ultimately, it fosters im-proved employability, structured career progression, and a scalable ecosystem that supports data-driven decision-making in mentorship and internship management.
InsightIQ: Redefining Truth In The Digital Era
Authors: Mr.M.Thangadurai, Sabari.M, Santhosh kumar.J, Sheshu Aakaash P
Abstract: In an age of information overload and digital misinformation, InsightIQ emerges as a transforma-tive solution for discerning truth from falsehood. This innovative platform leverages advanced artificial intelligence, natural language processing, and machine learning algorithms to analyze, verify, and contextualize information across digital ecosystems. InsightIQ employs multi-source cross-referencing, credibility scoring, and real-time fact-checking to combat fake news, deep-fakes, and manipulated content. The system integrates semantic analysis with blockchain-verified data trails, ensuring transparency and accountability in information dissemination. By empower-ing users with evidence-based insights and detecting logical fallacies, InsightIQ fosters critical thinking and digital literacy. The platform addresses contemporary challenges including echo chambers, algorithmic bias, and coordinated disinformation campaigns. Through its adaptive learning framework, InsightIQ continuously evolves to identify emerging manipulation tech-niques while maintaining ethical standards and user privacy. This groundbreaking tool represents a paradigm shift in how society consumes, evaluates, and trusts digital information, ultimately strengthening democratic discourse and informed decision-making in our interconnected world.
InsightIQ: Redefining Truth In The Digital Era
Authors: Mr.M.Thangadurai, Sabari.M, Santhosh kumar.J, Sheshu Aakaash P
Abstract: In an age of information overload and digital misinformation, InsightIQ emerges as a transforma-tive solution for discerning truth from falsehood. This innovative platform leverages advanced artificial intelligence, natural language processing, and machine learning algorithms to analyze, verify, and contextualize information across digital ecosystems. InsightIQ employs multi-source cross-referencing, credibility scoring, and real-time fact-checking to combat fake news, deep-fakes, and manipulated content. The system integrates semantic analysis with blockchain-verified data trails, ensuring transparency and accountability in information dissemination. By empower-ing users with evidence-based insights and detecting logical fallacies, InsightIQ fosters critical thinking and digital literacy. The platform addresses contemporary challenges including echo chambers, algorithmic bias, and coordinated disinformation campaigns. Through its adaptive learning framework, InsightIQ continuously evolves to identify emerging manipulation tech-niques while maintaining ethical standards and user privacy. This groundbreaking tool represents a paradigm shift in how society consumes, evaluates, and trusts digital information, ultimately strengthening democratic discourse and informed decision-making in our interconnected world.
Artificial Intelligence Blockchain Based Fake News Discrimination
Authors: I.Sravani, N. Nageswari, M. Sravanthi, P. Chandrasekhar, S.Sameer Hussain
Abstract: This paper minimizes fake news, which has been a hot topic recently, using blockchain and artifi-cial intelligence technology, and verifies it with blockchain. Also, using Artificial Intelligence technology, we want to create an algorithm that predicts how fake news will spread in the future. You can see various attempts at a news media platform based on Blockchain technology. Howev-er, the Blockchain news media platform is still not getting the market response we expected. It is questionable whether the reason is simply because it is a new technology, so it takes a long time to gain trust from consumers, whether consumers are not yet expecting an innovative news media platform, or whether the explosive growth of the Blockchain news media platform is difficult for other reasons. Research to answer this or direct research between Blockchain and media plat-forms is still lacking. In addition, the method of verifying fake news using artificial intelligence was verified, ANN, CBR, and MDA were changed, and the experiment was verified for pro-gress. In addition, the use of 5-fold cross-validation as a comparative method was added as de-scribed above to more closely examine the possibility of its usefulness even in general situations. Also, through various fields of artificial intelligence and blockchain, verification work was done with blockchain, and fake news prediction was made using artificial intelligence. Various experi-ments were conducted and performance tests were performed, while the performance of about 5,000 TTPS was recorded through the third experiment. In the future, we think it is necessary to combine Artificial Intelligence and blockchain technology.
Improving Sleep Disorder Diagnosis Through Optimized Machine Learning Approaches
Authors: Dr.Y Subba Reddy, B. Veera Vardhan, G. Sri Charan, H. Mohammad Abbas, K. Mohammad Jafar
Abstract: Sleep disorders, including Sleep Apnea and Insomnia, significantly affect individuals' health and quality of life, necessitating accurate and accessible diagnostic methods. Traditional diagnostic tools, such as Polysomnography (PSG), are expensive, time-consuming, and limited in accessi-bility, often leading to delayed or missed diagnoses. This project aims to address these limitations by leveraging machine learning algorithms for the classification of sleep disorders using the Sleep Health and Lifestyle Dataset. The existing system utilizes traditional algorithms such as K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Decision Tree, Random Forest, and Artifi-cial Neural Network (ANN). However, these approaches face challenges like computational overhead, sensitivity to hyperparameters, and limited interpretability. To overcome these issues, the proposed system implements advanced ensemble learning techniques, including the Stacking Classifier and Voting Classifier, to improve accuracy, robustness, and scalability. The project comprises data preprocessing, feature engineering, and model training using health and lifestyle features such as sleep duration, quality of sleep, physical activity, and stress levels. The system also provides users with an intuitive interface to upload data, view predictions, and analyze re-sults. Additionally, it visualizes the distribution of sleep disorder types to enhance diagnostic understanding.
Early Prediction Of Parkinson’s Disease Using Machine Learning
Authors: Dr.D.Siva Sankar Reddy, C.Tripurambika, G.Vyshnavi, G.prasanna Lakshmi, K.Sai Kiran Kumar
Abstract: Parkinson’s Disease (PD) is a neurodegenerative disorder that affects movement control, leading to symptoms such as tremors, stiffness, and bradykinesia. Early detection plays a crucial role in managing the disease effectively. Traditional diagnostic methods often require medical imaging or clinical assessments, which can be time-consuming and expensive. This project explores the use of machine learning models to predict Parkinson’s Disease from speech data, a non-invasive and accessible source. By analyzing features such as pitch, tone, and rhythm from speech samples, the project leverages machine learning algorithms like Support Vector Machine (SVM), K-Nearest Neighbors (KNN) to classify whether an individual exhibits signs of Parkinson’s Disease. Addi-tionally, the project incorporates. The developed system provides an intuitive interface where users can upload speech samples and receive predictions, offering a potential tool for early Par-kinson’s Disease detection and aiding healthcare professionals in diagnosis.
DOI:
Design And Performance Analysis Of Hybrid Solar Assisted Heat Exchanger For Sustainable Thermal Energy Applications
Authors: Mr. K. L. Kumar, Dr. G. K. Manikandan, Mr. M. Nandhakumar, Prof. R. Pandiyarajan
Abstract: This paper provides an extensive analysis of hybrid solar-assisted heat exchangers, with special emphasis on design innovations and performance optimization techniques for sustainable thermal energy systems. In this study, the researcher has conducted an exhaustive analysis of recent ex-perimental and computational investigations conducted between 2021 and 2026 to evaluate the important design factors in solar-assisted heat exchangers, including fin designs, phase change material, nanofluid utilization, and machine learning-based optimization techniques. An Integrated Solar Hybrid Heat Exchanger Performance Framework (ISHHEPF) has been proposed to assess the thermal efficiency, electrical efficiency, energy storage capacity, and overall performance of hybrid heat exchangers. From the analysis, it has been concluded that fin-type heat exchangers exhibit superior thermal performance, with prototypes showing heat transfer coefficient values up to 5790 W/m²°C and fluid outlet temperatures above 75°C at standard operating conditions. In addition, the combination of PCMs with nanofluids has been found to improve thermal storage capacity, with coconut oil-based PCMs showing 220.4 kJ/kg energy storage capacity with 67.1% thermal efficiency. Machine learning optimization techniques, such as XGBoost with the applica-tion of metaheuristic algorithms, show an improvement of 15% in thermal efficiency and an in-crease of 27% in exergy efficiency. The application of the techniques in the industry shows lev-elized cost of heat reductions of up to 54% compared to individual alternatives. The comparative evaluation of the alternatives in six analytical dimensions—design configuration, thermal perfor-mance, electrical efficiency, energy storage, economic viability, and optimization methodology—indicates that the hybrid PV/T system with advanced heat transfer enhancements is a promising alternative for sustainable thermal energy.
Prediction of Air Pollution Using Machine Learning Models
Authors: Dr. Pandurangan Ravi, B. Raghunath Reddy, A. Rajasekhar Reddy
Abstract: Air pollution has become one of the most critical environmental challenges worldwide, signifi-cantly impacting human health, climate change, and overall ecological balance. Rapid industriali-zation, urbanization, and increased vehicular emissions have led to a drastic rise in air pollutants such as particulate matter (PM2.5 and PM10), carbon monoxide (CO), nitrogen dioxide (NO₂), sulfur dioxide (SO₂), and ozone (O₃). Accurate prediction of air pollution levels is essential for implementing effective control measures, improving public awareness, and supporting policy-making decisions. This project focuses on the development of a machine learning-based predic-tive system to forecast air pollution levels using historical and real-time environmental data. The proposed system utilizes various machine learning algorithms such as Linear Regression, Deci-sion Tree, Random Forest, Support Vector Machine (SVM), and Artificial Neural Networks (ANN) to analyze patterns and relationships among different pollution indicators and meteorolog-ical parameters including temperature, humidity, wind speed, and atmospheric pressure. The dataset used in this study is collected from reliable sources such as government pollution control boards and environmental monitoring agencies. Data preprocessing techniques such as handling missing values, normalization, and feature selection are applied to improve model performance and accuracy. Exploratory Data Analysis (EDA) is conducted to identify trends, seasonal varia-tions, and correlations between pollutants and weather conditions.
AI Based Air Pollution Monitoring And Prediction System
Authors: Ms.K. Madhumitha, M Mohana Priya, R Priya, D Namitha
Abstract: Air pollution is one of the most serious environmental and health concerns in modern urban societies. This study proposes an AI-driven Air Pollution Monitoring and Prediction System that enables continuous observation and accurate forecasting of air quality levels. The system employs IoT-based sensors to gather real-time data on pollutants such as PM2.5, PM10, CO, NO₂, and SO₂, along with meteorological parameters including temperature and humidity. The collected data is transmitted to a cloud environment for preprocessing and analysis. Advanced machine learning and deep learning techniques, particularly time-series models like LSTM and regression algorithms, are utilized to predict future Air Quality Index (AQI) levels. The system also incorpo-rates interactive dashboards and automated alert mechanisms to inform authorities and the public about potential pollution risks. Performance evaluation indicates higher prediction accuracy and lower error margins compared to conventional statistical approaches. The proposed framework supports data-driven environmental management, timely intervention strategies, and sustainable urban development while promoting improved public health outcomes.
Student Freelancing Job Portal
Authors: Lalitha K, Ashley Jude A, Jawahar T, Mohamed Ibrahim T
Abstract: Student Freelancing Job Portal is a digital platform designed to connect students with freelance opportunities that match their skills, interests, and academic schedules. The system enables stu-dents to create profiles, showcase portfolios, and bid on projects posted by clients across domains such as content writing, graphic design, programming, and data entry. It addresses the challenges students face in finding flexible, part-time work by providing a centralized, user-friendly interface with secure payment integration and transparent rating mechanisms. Employers benefit from access to a diverse, affordable talent pool, while students gain real-world experience, income, and professional exposure. The portal incorporates features like skill verification, job recommenda-tions using basic algorithms, communication tools, and feedback systems to ensure quality and trust. Additionally, it promotes career readiness by helping students build networks and improve employability. Overall, the platform bridges the gap between education and industry by fostering a mutually beneficial ecosystem for students and recruiters in the growing gig economy. It also includes mobile accessibility, multilingual support, and analytics dashboards for tracking perfor-mance, earnings, and engagement, ensuring continuous improvement and scalability of the plat-form effectively.
Enterprise Modernization Through Hybrid Cloud Deployment: Evidence Mapping Of Multi-Environment Strategies
Authors: Dr. Ajay Varma Indukuri, Dr. Rajeev Samuel Devadas, Dr. Prashanth Reddy Kora, Dr. Venkata Raghavendra Vutti, Chaitanya Srinivas
Abstract: Enterprise modernization has become a critical priority for organizations seeking to improve scalability, agility, and operational efficiency in an increasingly digital environment. Hybrid cloud deployment has emerged as a strategic approach that integrates on-premises infrastructure with public and private cloud services, enabling enterprises to leverage the benefits of multiple compu-ting environments. This study presents an evidence mapping analysis of multi-environment strat-egies used in hybrid cloud deployments to support enterprise modernization initiatives. The re-search synthesizes existing literature and industry practices to identify key deployment models, integration mechanisms, workload distribution approaches, and governance frameworks adopted across diverse enterprise settings. The findings highlight how multi-environment strategies en-hance flexibility, optimize resource utilization, improve security compliance, and support gradual migration from legacy systems to modern cloud-native architectures. Additionally, the evidence mapping reveals common challenges such as interoperability, data consistency, management complexity, and security integration across heterogeneous environments. By systematically cate-gorizing and analyzing available evidence, this study provides a structured overview of current hybrid cloud deployment practices and offers insights that can guide organizations in designing effective multi-environment strategies for successful enterprise modernization.
A Study On Policy-Driven Automation For Efficient Enterprise Data Platform Management
Authors: Dr. Jonathan R. Miller, Dr. Emily K. Thompson, Matthew S. Collins, Chaitanya Srinivas
Abstract: Policy-driven automation has emerged as a critical enabler for managing the growing complexity of enterprise data platforms in the era of digital transformation. This study examines the design, implementation, and impact of policy-driven automation in enhancing the efficiency, scalability, and governance of enterprise data environments. By integrating rule-based policies with automat-ed workflows, organizations can streamline data operations, ensure compliance with regulatory standards, and reduce manual intervention. The research explores key components such as policy definition, orchestration mechanisms, and real-time monitoring, highlighting their role in optimiz-ing resource utilization and improving system reliability. Furthermore, the study evaluates chal-lenges including policy conflicts, integration with heterogeneous systems, and maintaining adapt-ability in dynamic environments. The findings demonstrate that policy-driven automation not only accelerates data processing and decision-making but also strengthens data governance frame-works, making it a vital approach for modern enterprise data platform management.
Personalized Career Skill Development System
Authors: Mrs.T. Dheepa, Dharshini S, Haripriya A, Madhumitha M
Abstract: In today’s competitive job market, students from different educational backgrounds often struggle to understand whether their current skills align with their desired career paths. This project presents a Personalized Career Skill Development System, a mobile application designed to help learnersfromArts,Science,Engineering,Commer ce,andotherstreamsassesstheirskillreadinessand plan effective career development. The system allows users to input their existing skills, study level, and preferred career stream. Based on this information, the application evaluates the identifiesskillgaps,andprovidesapersonalizedlea rningroadmaptailoredtotheselectedcareer path. Instead of only highlighting missing skills, the app activelym supports skill improvement through learning modules, practice exercises, assessments, and mini projects. The application also tracks user progress, suggests suitable job roles, and focuses on improving weak skill areas to ensure continuous development. By offering structured guidance and personalized learning support, the system helps users become better prepared for their chosen careers. This solution aims to bridge the gap betweenedu-cationandemployability,makingitav aluabletoolforstudentsandearly- careerlearnersacrossall do-mains.
Comparative Investigation Of Lophira Alata Sawdust And Activated Carbonized Sawdust In Treating Of Heavy-Metal Contaminated Water
Authors: Oluchukwu Benedicta Chikwe, Ikechukwu Sampson Chikwe, Onwugbuta Godpower Chukwuemeka, Ogbonnaya Mba Arunsi , Stanley Sotonye, Korie Maximus Chibuoyi, Eresanya Olanrewaju Isola, Erienu Obruche Kennedy
Abstract: The Lophira alata wood sawdust obtained for this study was first thoroughly washed with distilled water to remove impurities, then dried under controlled conditions, and subsequently divided into two equal portions. The first portion was retained and used directly as the unmodified sawdust sample. The second portion underwent further treatment: it was carbonized at a temperature of 600 °C for 4 hours to enhance its structural properties, and then chemically activated using 2 M KOH for 24 hours at room temperature to improve its adsorption capacity. Both prepared samples were utilized as adsorbents for the removal of Nickel (Ni²⁺) ions from aqueous solutions. Key operational parameters affecting adsorption, including contact time, adsorbent dosage, and pH, were systematically investigated. Additionally, the physicochemical characteristics of both adsorbents were evaluated to understand their performance. The results revealed that increasing the sawdust dosage, contact time, and solution pH significantly enhanced the adsorption efficiency. Moreover, the activated-carbonized sawdust demonstrated a higher adsorption capacity compared to the unmodified sample. Therefore, Lophira alata sawdust, particularly in its activated-carbonized form, is an effective, low-cost material for removing toxic heavy metals from wastewater.
An AI Assisted Web Tool For Survey Data Cleaning And Statistical Reporting
Authors: Dr. J. Yogapriya, Abinithi S, Dhivya V, Harshavarthini V
Abstract: Survey-based research often faces data quality issues such as missing values, duplicate entries, inconsistent responses, and outliers, which can affect the accuracy of statistical analysis and deci-sion-making. This paper proposes an AI- assisted web-based system for automated survey data cleaning and statistical reporting. The system integrates rule-based techniques and Large Lan-guage Model (LLM) capabilities to detect missing values, remove duplicates, identify outliers, and standardize survey responses with minimal manual intervention. After preprocessing, the platform automatically generates descriptive statistical reports that help researchers quickly ana-lyze cleaned datasets. The proposed system improves the reliability and efficiency of survey data analysis while reducing manual preprocessing effort. It supports applications in healthcare sur-veys, educational research, and workforce studies, contributing to data-driven decision-making aligned with Sustainable Development Goals (SDGs) including Good Health and Well-Being, Quality Education, Decent Work and Economic Growth, Indutry Innovation and Infrastructure, and Partnerships for the Goals.
DOI: https://doi.org/10.5281/zenodo.19386018
An IoT- Based Automatic Fish Feeder
Authors: Mrs. T. Thenmozhi, Deepika E, Janani M, Lakxitha N
Abstract: Fish owners and small-scale aquaculture systems often face difficulties in maintaining regular feeding schedules due to busy routines or absence from home. Irregular feeding can affect fish health, growth, and water quality. This project presents an IoT-based Automatic Fish Feeder for smart and remote fish feeding, a web-based system that allows users to control feeding operations through a Wi-Fi connected interface. The system is developed using an ESP32-CAM microcon-troller integrated with a servo motor attached to a food container to dispense a controlled amount of fish feed. When a feeding command is given through the web interface, the servo motor rotates and releases a measured quantity of food into the tank. The ESP32-CAM module also provides a live camera view, enabling users to monitor the feeding process and confirm that feeding has occurred successfully. The system operates through wireless connectivity without requiring com-plex additional hardware, making it suitable for home aquariums and small-scale fish farming setups. By enabling timely feeding and remote monitoring, the system improves fish care while reducing manual effort. This approach supports Sustainable Development Goal 2 (Zero Hunger), Sustainable Development Goal 9 (Industry, Innovation and Infrastructure), and Sustainable De-velopment Goal 12 (Responsible Consumption and Production) by promoting efficient and re-sponsible feed management in aquaculture systems.
DOI: https://doi.org/10.5281/zenodo.19386226
AI Powered Deepfake Detection For Secure Video Conferencing
Authors: K.Anguraju, V Mohan, M.Santhosh Sivan, N.Vishnu Prasad
Abstract: The rapid advancement of artificial intelligence has led to the emergence of deepfake technology, which poses significant threats to the security and authenticity of video conferencing systems. This paper proposes an AI-powered deepfake detection framework designed to ensure secure and trustworthy virtual communication. The system utilizes deep learning techniques, including con-volutional neural networks (CNNs) and recurrent neural networks (RNNs), to analyze facial expressions, lip synchronization, and temporal inconsistencies in video streams. By extracting both spatial and temporal features, the model effectively distinguishes between genuine and ma-nipulated video content in real time. Additionally, the framework integrates anomaly detection and metadata analysis to enhance detection accuracy. Experimental results demonstrate that the pro-posed system achieves high precision and recall while maintaining low latency, making it suitable for real-time deployment in video conferencing platforms. This approach strengthens cybersecuri-ty measures and helps prevent identity fraud, misinformation, and unauthorized access during online meetings.
DOI:
COGNIX – A Cognitive AI Agent Platform For Real-Time Business Intelligence And Automation
Authors: Dr. J. Yogapriya , Ravichandran P, Naveeth R P, Thiruvasan s
Abstract: The increasing digitization of healthcare systems has led to the widespread adoption of electronic health records (EHRs), raising concerns about data security, privacy, and interoperability. This paper proposes a blockchain-based secure health record management system designed to ensure the confidentiality, integrity, and accessibility of patient data. By leveraging the decentralized and tamper-resistant nature of blockchain technology, the system enables secure storage and sharing of medical records among authorized stakeholders such as hospitals, doctors, and patients. Smart contracts are utilized to enforce access control policies and automate data-sharing permissions. Additionally, cryptographic techniques are integrated to protect sensitive information from unau-thorized access and cyber threats. The proposed framework enhances transparency, reduces the risk of data breaches, and eliminates the need for a centralized authority. Experimental analysis demonstrates improved data security and efficient record management compared to traditional systems. This approach offers a reliable and scalable solution for modern healthcare environ-ments.
DOI: https://doi.org/10.5281/zenodo.19471793
Smart Virtual Companion With Personalized Conversations
Authors: R.Premkumar, Janesh M, Gowsik PU, Joachim Andrew A
Abstract: The growing need for intelligent and interactive digital assistants has driven the development of smart virtual companions that offer highly personalized user experiences. This paper introduces a Smart Virtual Companion with Personalized Conversations, aimed at improving human-computer interaction through adaptive and context-aware dialogue. The system utilizes advanced natural language processing (NLP) and machine learning methods to interpret user intent, preferences, and emotional context. By storing user profiles and analyzing past interactions, the companion is able to generate tailored responses that enhance engagement and overall user satisfaction.The proposed framework incorporates key components such as sentiment analysis, intent detection, and efficient dialogue management to enable meaningful and dynamic conversations. It continu-ously learns from ongoing interactions, allowing it to refine its responses and better align with user needs over time. The system is designed for deployment across various platforms, including mobile devices, web interfaces, and smart home systems, ensuring flexibility and accessibility. Furthermore, built-in privacy measures safeguard user data during interactions. Experimental evaluations indicate that the system outperforms traditional rule-based chatbots in terms of con-versational quality and user engagement.
DOI: https://doi.org/10.5281/zenodo.19471913
Blockchain-based Secure Health Record Management
Authors: V Dhanalakshmi, Prabhadevi R, rivarshini M, Vaishnavi S
Abstract: The increasing digitization of healthcare systems has led to the widespread adoption of electronic health records (EHRs), raising concerns about data security, privacy, and interop-erability. This paper proposes a blockchain-based secure health record management system designed to ensure the confidentiality, integrity, and accessibility of patient data. By lever-aging the decentralized and tamper-resistant nature of blockchain technology, the system enables secure storage and sharing of medical records among authorized stakeholders such as hospitals, doctors, and patients. Smart contracts are utilized to enforce access control policies and automate data-sharing permissions. Additionally, cryptographic techniques are integrated to protect sensitive information from unauthorized access and cyber threats. The proposed framework enhances transparency, reduces the risk of data breaches, and elimi-nates the need for a centralized authority. Experimental analysis demonstrates improved data security and efficient record management compared to traditional systems. This ap-proach offers a reliable and scalable solution for modern healthcare environments.
DOI: https://doi.org/10.5281/zenodo.19474743
Remediation Of Surface Water Polluted By Mining Effluent Discharges: Environmental Impacts And Treatment Methods In South-South Region, Nigeria
Authors: Ikechukwu S. Chikwe, Odioko M. Obianuju, Onadje F. Ovwighose, Apuyor K. Efe, Onwugbuta G. Chukwuemeka, Chikwe Oluchukwu B, Anyanwu Gogo.C, Stanley Sotonye, Etus Patrick Chimuanya, Joseph N. Bar
Abstract: Heavy metals, including lead (Pb) and mercury (Hg), are highly toxic to both the environment and human health. Prolonged exposure to these metals can lead to significant health disorders. This study aimed to find an effective adsorbent to reduce Pb and Hg concentrations in water samples from South-South Region, Nigeria. The river was found to be highly turbid (average of 355 NTU), making it unsuitable for domestic use without treatment. The research tested modified rice husk (RH-TAM) and orange peels (OP-TAM) using tartaric acid. Results showed that the modified rice husk (RH-TAM) exhibited superior adsorption efficiency for Pb and Hg. Batch experiments were conducted to evaluate the removal efficiency of Pb and Hg using various ad-sorbents (modified and unmodified rice husk and orange peel). Factors such as pH, contact time, and adsorbent dosage were important in the sorption process. The optimal conditions were found to be pH 5, an adsorbent dosage of 0.5 g/20 ml, and a contact time of 4 hours at 35°C. Under these conditions, the highest Pb adsorption efficiencies were 75.56% for RH- TAM and 69.93% for unmodified rice husk (UM-RH). For Hg, RH-TAM achieved 53.26%, while UM-RH reached 45.11%. The adsorption efficiency of OP-TAM was 62.03% for Pb and 44.57% for Hg, with unmodified orange peel (UM-OP) showing the lowest efficiencies. The Langmuir isotherm better fitted the experimental data for both metals.
Wireless Power And Data Link For Ev
Authors: Tejaswini Patil, Parth Nikam, Niranjan Kulkarni, Sarthak Jadhav, Mr.A.K.Sonawane
Abstract: The rapid adoption of electric vehicles (evs) has intensified the demand for efficient, safe, and user-friendly charging technologies. Conventional plug-in charging systems present limitations related to physical wear, safety risks, and user inconvenience. This paper presents a comprehen-sive study of wireless power and data transfer systems for EV applications, focusing on the integration of inductive power transfer (IPT) with real-time communication mechanisms. The proposed approach enables simultaneous energy transmission and bidirectional data exchange between the ground infrastructure and the vehicle, ensuring optimized charging control, alignment detection, and system monitoring. Key design considerations such as coupling efficiency, misa-lignment tolerance, electromagnetic compatibility, and communication reliability are analyzed. Furthermore, the paper explores system architectures, compensation techniques, and control strategies to enhance overall performance. Simulation and experimental insights demonstrate that the integrated wireless system achieves high efficiency while maintaining robust data communica-tion. The proposed framework contributes to the advancement of autonomous and intelligent charging infrastructure, supporting the future development of smart transportation systems.
Intelligent Fault Prediction In Cloud Systems Using Deep Learning Techniques
Authors: S.Jayashree Ananth, Mr Naveen VS
Abstract: Due to their evolving nature and growing complexity, cloud computing systems require efficient fault management techniques that would assure availability and reliability of provided services. This paper introduces a deep learning-based fault prediction framework designed specifically for clouds. It utilizes a combination of Bidirectional Gated Recurrent Units (Bi-GRU), attentional mechanisms, and Graph Neural Networks (GNNs). The proposed model incorporates temporal dependency from cloud system telemetry data and also accounts for dependencies between microservices within the system. Testing on real-life datasets such as Google Cluster Trace and Alibaba Cluster data showed 96.2% prediction accuracy, 92.8% precision and 91.5% recall, which outperforms current fault prediction techniques by 8-12%. Additionally, due to its attention-based architecture, the model is capable of providing explainability by highlighting important temporal parts and specific services at risk. Results show that the proposed approach allows for implementing proactive fault prevention techniques that reduce SLA violation rate by 65% and cut down recovery times by 55%.
DOI: https://doi.org/10.5281/zenodo.19698844
AIoT-Enabled Digital Twin Systems For Industrial Automation And Smart Manufacturing
Authors: Nafisa S, Dr. Balaji. K
Abstract: The integration of AIoT technology with DT technologies is the basis of a new industrial revolution in industrial automation and intelligent manufacturing. This study develops a framework for AIoT-based digital twin systems, which combines live IoT data with AI simulation and optimization models. The designed model is based on a four-layer cyber-physical structure including data gathering from the edge, stochastic simulation, state encoding using graph attention networks, and closed-loop execution. The framework was analyzed using 10,000 stochastic simulations and a 12-week industrial experiment in which the system performed schedule performance of 96.8%, OEE of 84.7%, and 16.5% reduction in energy consumption per tonnage produced. The developed multi-objective reinforcement learning algorithm showed an integrated relationship between waste reduction and increased OEE (r = -0.73), with a total OEE improvement of 34.1% due to sustainable processes. The global AI-powered digital twin market is forecasted to grow up to billion by 2030 with 26.2% CAGR.
DOI: https://doi.org/10.5281/zenodo.19699028
A Review On Explainable AI-Based Risk Prediction Model For PCOS Diagnosis Using Machine Learning
Authors: Er. Mamta Bhardwaj, Dr. Komal Garg
Abstract: Polycystic Ovary Syndrome (PCOS) is a complex endocrine disorder-affecting women of reproductive age, characterized by hormonal imbalance, metabolic complications, and reproductive issues. Early diagnosis remains challenging due to heterogeneous symptoms and reliance on subjective clinical criteria. Recent advancements in Machine Learning (ML) have shown promising results in improving diagnostic accuracy; however, the lack of interpretability limits their adoption in clinical practice. This review paper presents a comprehensive analysis of ML-based PCOS prediction models with a focus on Explainable Artificial Intelligence (XAI). It explores the role of ML algorithms such as Logistic Regression, Decision Trees, Random Forest, Support Vector Machines (SVM), and ensemble techniques in enhancing prediction performance. Additionally, the study highlights preprocessing methods including SMOTE, feature scaling, and feature selection techniques that improve model efficiency. This review provides valuable insights into the development of accurate, interpretable, and reliable PCOS diagnostic systems, bridging the gap between computational intelligence and clinical applicability.
Energy Management System On Multi-Microgrid System Using Metaheuristic Algorithms
Authors: Dr.M. Sri Suresh, A. Sai Pranav Reddy, B. Prem Kumar, P. Ganesh
Abstract: The escalating environmental pollution have become major issues that require novel and eco-friendly methods. Integration of renewable energy is one of the possible solutions. resources (RERs) and effective energy. management strategies. Energy management aims at minimizing. and as well as operating, maintenance and generation costs. enhancing system performance by means of methods like minimization of power losses, stability improvement, and emission. reduction. In this respect, the energy management of mi-crogrids has played out to be a major concern in the modern. power systems. An optimi-zation is presented in this paper. model of a multi-objective problem of a renewable. Mul-ti-microgrid (MMG) system is based on energy. The system consisting of three intercon-nected microgrids, all equipped. and wind turbines (WT) and photovoltaic (PV) panels, became part of the IEEE 33-bus distribution system. The model takes into consideration the variation in PV and WT output, load Sporadic demand, and real-time prices of elec-tricity. Three objective functions are designed in a way that they reduce the overall. cost/year, deviation of voltage, and voltage stability index- developing a cost-performance multi-objective collectively. optimization problem. With the assistance of the, the issue is considered. Particle Swarm Optimization (PSO) algorithm, both with and and without RERs.. Additionally, a comparative analysis is conducted using two other optimization techniques: Mountain Gazelle Optimization (MGO) and Gorilla Troop Optimization (GTO). Simulation results demonstrate that the proposed approach significantly reduces system costs and enhances overall performance.
A Study on Cybersecurity Challenges in Digital Marketing from the Perspective of Online Consumers in Coimbatore District
Authors: Assistant Professor Ms.R.Nandhini, Ms.Priyanka
Abstract: In the modern digital era, digital marketing has become a vital tool for businesses to reach and engage customers effectively. However, with the increasing dependence on digital platforms, cybersecurity challenges have also emerged as a major concern for online con-sumers. This study focuses on identifying and analysing cybersecurity issues faced by con-sumers in Coimbatore district while engaging in digital marketing activities such as online shopping, social media interactions, and digital transactions. The research highlights key concerns such as data breaches, phishing attacks, identity theft, and lack of awareness re-garding online safety. The study also examines the level of consumer trust in digital plat-forms and their perception of security measures adopted by businesses. The findings reveal that while digital marketing offers convenience and accessibility, cybersecurity threats sig-nificantly affect consumer confidence. The study suggests strategies to enhance cybersecu-rity awareness and strengthen protective mechanisms to ensure safer digital experiences.
A Study on Stock Market Awareness Among Students with Special Reference to Coimbatore Districta Study on Stock Market Awareness Among Students with Special Reference to Coimbatore District
Authors: Assistant Professor Ms.V.Vineetha, Mr. Raja A
Abstract: This study examines the level of awareness of stock market investment among students. It focuses on their knowledge, risk perception, and factors influencing their investment deci-sions. Although students show interest in stock market activities, many lack sufficient knowledge and practical experience. The study uses primary data collected through ques-tionnaires and applies a descriptive research design. The findings reveal that limited finan-cial literacy and fear of risk act as major barriers. The study suggests that improving finan-cial education can help students make informed investment decisions.
A Study on Customer Satisfaction of Mobile Banking with Special Reference to Coimbatore City
Authors: Assistant Professor Ms. Swathi .M, Mr. Vignesh V
Abstract: This study examines customer satisfaction with mobile banking services, with special refer-ence to Coimbatore City. With the rapid growth of digital technology and increased smartphone penetration, mobile banking has become an essential service in the banking sector. The study aims to analyze the level of customer satisfaction, identify the factors influencing user experience, and understand the challenges faced by customers while using mobile banking applications. Primary data was collected through structured questionnaires from mobile banking users in Coimbatore City, and appropriate statistical tools were used for analysis. The findings reveal that convenience, security, ease of use, and service quality significantly influence customer satisfaction. However, issues such as technical glitches, network problems, and security concerns continue to affect user experience. The study suggests that banks should enhance app performance, strengthen security features, and provide better customer support to improve overall satisfaction. This research contributes to understanding customer perceptions and offers insights for improving mobile banking ser-vices in urban areas.
A Study on Market Analysis of FMCG Products in Urban Areas with Special Reference to Coimbatore
Authors: Assistant Professor Ms. Revathi.G, Mr. Yanish Krishna.S
Abstract: The Fast-Moving Consumer Goods (FMCG) industry is a crucial sector in the Indian econ-omy, comprising products such as packaged foods, beverages, toiletries, over-the-counter medicines, and household essentials that are sold quickly at relatively low prices. This study examines the FMCG market in Coimbatore's urban areas by analysing consumer prefer-ences, competitive forces, and marketing strategies. It highlights the impact of digital trans-formation, lifestyle changes, and demographic shifts on consumer behaviour, helping busi-nesses adapt and sustain growth. The study is based on primary data collected from 120 urban respondents and employs percentage analysis and chi-square testing. Findings reveal that brand loyalty, price sensitivity, and product availability are the dominant purchase decision factors. The research provides actionable insights for FMCG marketers operating in Tier-II Indian cities.
A Study on Risk and Return Analysis of Mutual Funds
Authors: Assistant Professor Mr.P.Sasikumar, Mr. R. Periyasamy
Abstract: This study focuses on analyzing the risk and return characteristics of mutual funds, which have become a popular investment avenue among individuals. The research evaluates different types of mutual funds based on their performance, volatility, and return patterns to understand how effec-tively they balance risk and reward. By using statistical tools and comparative analysis, the study highlights the relationship between risk levels and expected returns. The findings aim to assist investors in making informed decisions and contribute to a better understanding of mutual fund performance in the financial market.
Awareness of Systematic Investment Plans (Sips) Among Young Investors in Coimbatore
Authors: Assistant Professor Ms. Mithuna R, Ms. Pooja Sri C
Abstract: Systematic Investment Plans (SIPs) have become one of the most popular investment options in India for individuals seeking disciplined and long-term wealth creation. SIPs allow investors to invest a fixed amount regularly in mutual funds, helping them benefit from rupee cost averaging and the power of compounding. Despite the rapid growth of the mutual fund industry and in-creasing financial awareness, many young investors still lack sufficient knowledge about SIPs and their benefits. This study focuses on analysing the awareness of Systematic Investment Plans among young investors in Coimbatore city. The research aims to understand the level of knowledge, demographic factors influencing awareness, and the investment behaviour of individ-uals aged between 18 and 35 years. Primary data was collected using a structured questionnaire from a sample of respondents in Coimbatore. Secondary data was obtained from journals, books, and financial reports. The data collected from respondents was analysed using statistical tools such as Chi-Square analysis to examine the relationship between demographic variables and SIP awareness. The findings reveal that factors such as education level, income, and occupation sig-nificantly influence the awareness and adoption of SIP investments among young investors. The study concludes that although awareness of SIPs is increasing, many young individuals still require financial education regarding mutual fund investments. The research suggests that finan-cial institutions, educational institutions, and regulatory bodies should undertake awareness pro-grammes to improve financial literacy and promote SIP investments among youth.
A Study on Digital Checkout Friction and Cart Abandonment Among E-Commerce Customers in Coimbatore City
Authors: Assistant Professor Mr.M.Eknath Prasath, Mr.Sakthisanjith.A
Abstract: Digital technology has significantly transformed the way consumers shop, especially through e-commerce platforms. This study examines digital checkout friction and cart abandonment behav-iour among e-commerce customers, with special reference to Coimbatore City. Checkout friction, such as extra charges, lengthy processes, payment issues, and security concerns, plays a major role in influencing purchase decisions. The study aims to understand customer behaviour, identify the key reasons for cart abandonment, and analyse factors affecting checkout completion. The study is based on primary data collected through a structured questionnaire from respondents. The findings reveal that while many customers frequently shop online, a significant number abandon their carts due to factors like additional costs, slow website performance, and lack of preferred payment options. Security concerns and complicated checkout processes also contribute to incomplete transactions.Overall, the study highlights the importance of improving website design, ensuring transparency in pricing, enhancing payment security, and simplifying checkout procedures. These measures can help reduce cart abandonment and improve customer satisfaction in e-commerce platforms.
A Study on the E-Commerce Platforms and Their Impact on Rural Micro and Small Enterprises With Special Reference to Coimbatore District
Authors: Assistant Professor Dr. G. Kowsalya, Ms. A. Shridharshini
Abstract: This study examines the impact of e-commerce platforms on rural micro and small enterprises in Coimbatore district. The growth of digital technology has enabled rural entrepreneurs to access wider markets through platforms like Amazon, Flipkart, and Meesho. Data was collected from 90 respondents, including business owners, artisans, and traders, using structured questionnaires. The analysis was conducted using percentage methods and chi-square tests. Findings show that e-commerce helps increase sales, expand market reach, and improve income levels. However, challenges such as lack of digital literacy, logistics issues, and limited awareness persist. These barriers restrict the full utilization of online platforms. The study suggests enhancing digital train-ing and improving infrastructure. This would help rural entrepreneurs benefit more effectively from e-commerce opportunities.
A Study on Awareness on Share Trading Among College Students with Special Reference to Coim-batore City
Authors: Assistant Professor Ms.V.Priyanka, Mr. K.Sumesh
Abstract: This study explores the level of awareness about share trading among college students in Coim-batore City. With the growing significance of equity markets in India's financial ecosystem and the increasing accessibility of trading platforms through smartphones, it has become essential to understand the extent to which the younger generation particularly students is informed about and engaged with the stock market. The study investigates key dimensions including awareness levels, sources of information, trading experience, investment behavior, and the factors influenc-ing participation in share markets. Data was collected from 90 college students in Coimbatore City through structured questionnaires and analyzed using percentage analysis and chi-square tests. The findings indicate that while general awareness of share trading exists among students, actual participation remains low due to lack of knowledge, perceived risk, and financial con-straints. The study offers insights and recommendations for promoting financial literacy and responsible investment habits among youth.
A Study on Customer Satisfaction Towards Mobile Wallet Services with Reference to Paytm
Authors: Sri Vishnu, Assistant Professor Dr.G.Kowsalya
Abstract: This study examines customer satisfaction towards mobile wallet services with special reference to Paytm. With the rapid proliferation of smartphones and the expansion of digital payment infra-structure in India, mobile wallets have emerged as a convenient and widely-used mode of finan-cial transaction. Paytm, one of India's leading digital payment platforms, has transformed the way consumers manage money, make purchases, and transfer funds. The study investigates key di-mensions of customer satisfaction including ease of use, security, transaction speed, customer support, and service reliability. Data was collected from 90 respondents in Coimbatore district through structured questionnaires and analyzed using percentage analysis and chi-square tests. The findings reveal that while a majority of Paytm users are satisfied with transaction speed and ease of use, concerns persist around security, grievance redressal, and technical glitches. The study provides insights for digital payment service providers seeking
Blockchain-Enabled Secure Communication Framework For IoT Networks
Authors: Manikandan.R, Arivumani Samson S
Abstract: Due to the rapid expansion in the use of the Internet of Things (IoT), new challenges have emerged in relation to IoT security such as device verification, data integrity and fault toler-ance among others. This paper outlines a framework that provides secure communication in IoT using blockchain technology through a three layer design comprising of IoT devices, Blockchain consensus nodes and communication mechanisms. This proposed blockchain framework has a hybrid consensus scheme that combines the Practical Byzantine Fault Tolerance (PBFT) consensus mechanism with Proof-of-Authority (PoA). This scheme achieves a throughput improvement of 98.3% when compared to Ethereum based schemes with 95.3% lower latency. Evaluations conducted show transaction processing speeds of 1,080 transactions per second with a finality of 3.5 seconds supporting up to 400 devices concurrently. It is also able to support resilient operations with 33% Byzantine nodes and zero false positive rate for device authentication.
Dual-Input Solar PV–Battery EV Charging Station with INC-Fuzzy MPPT and Cascaded Sliding Mode Battery Control
Authors: Ms. D. Sushma, A. Prasanna Reddy, T. Hrishikesh, N. Sravani
Abstract: The accelerating shift toward electric mobility demands charging solutions that move beyond grid dependency and deliver consistent, renewable-backed power. This paper presents a dual-input solar photovoltaic (PV) and battery energy storage system (BESS) designed for EV charging, built around a two-phase interleaved boost converter. Two independently designed yet cooperatively operating control strategies govern the system: an Incremental Conductance–Fuzzy Logic hybrid MPPT (INC-Fuzzy) for the PV source, and a cascaded Sliding Mode Controller (SMC) for battery management. The INC-Fuzzy MPPT feeds a normalised INC error signal into a 49-rule Mamdani fuzzy inference engine, yielding adaptive variable-step tracking that surpasses conventional fixed-step methods in both convergence speed and steady-state accuracy. The SMC battery controller drives nested voltage and current sliding surfaces, generating dynamic charge and discharge duty signals while automatically transitioning between constant-current and constant-voltage modes based on battery state of charge (SOC). MATLAB/Simulink simulation results, presented as actual scope waveforms, confirm stable DC bus regulation at 324 V, smooth MPPT convergence, and well-controlled interleaved boost current from startup through steady state, demonstrating the practical viability of this architecture for sustainable EV charging infrastructure.
Impact Of Climatic Variables On Crop Yield Prediction Using Machine Learning Algorithms
Authors: Ambuj Kumar Misra
Abstract: The accelerating pace of climate change poses one of the most pressing challenges to global food security, making accurate and timely crop yield prediction an urgent scientific priority. This study investigates the influence of key climatic variables—including maximum and minimum temperature, precipitation, relative humidity, solar radiation, and atmospheric CO₂ concentration—on crop yield outcomes across diverse agricultural regions. Employing a suite of machine learning algorithms, namely Linear Regression, Support Vector Machines (SVM), Random Forest (RF), Gradient Boosting (GB), Long Short-Term Memory (LSTM) networks, and a hybrid CNN-LSTM architecture, we develop and evaluate predictive models using multi-decade observational data. Our findings demonstrate that ensemble methods and deep learning architectures substantially outperform traditional statistical models, with the CNN-LSTM hybrid achieving an R² score of 0.95 and a Root Mean Square Error (RMSE) of 0.14 t/ha. Precipitation and maximum temperature were identified as the most influential predictors. The results highlight the transformative potential of machine learning in enabling climate-adaptive agricultural planning and underscore the necessity of integrating climatic intelligence into yield forecasting systems.
Generative Ai Based Virtual Teaching Assistant For Personalized Learning
Authors: R Pavithra, Suhana F, Vishveshar B, Vijay Kumar K, Naveen, Vallarasu S
Abstract: The application of Generative Artificial Intelligence (GenAI) to education is set to transform personalized learning by offering adaptive, real-time tutoring according to the needs of learners. In this work, we outline an architecture for a GenAI virtual teaching assistant (VTA) based on large language models (LLMs), retrieval-augmented generation, knowledge tracing, and multi-modal content generation. The proposed GenAI-VTA architecture combines three main components: a knowledge tracing module based on deep knowledge tracing (DKT), an answer generation module powered by an LLM along with knowledge retrieval from a knowledge base aligned to the curriculum, and finally a learning analytics dashboard for teachers. Performance analysis based on a controlled experiment involving 150 undergraduate students reveals that the use of the GenAI-VTA increases learning performance by 28.7%, lowers average response time below 1.5 seconds, and provides satisfaction levels up to 86%.
Secure And Privacy-Preserving Federated Learning Framework For Environmental Sensing
Authors: Abhishek Dubey, Shivanshi Sahu, Shivansh Mishra, Tanvi Khetrapal, Sonali Patidar
Abstract: Environmental monitoring is essential for analyzing ecosystem behavior, forecasting natural hazards, and supporting sustainable resource use. With the rapid expansion of the Internet of Things (IoT) and distributed sensor networks, it has become possible to gather large-scale, real-time data. However, traditional centralized machine learning methods introduce significant chal-lenges related to data privacy, ownership, and regulatory requirements. Federated Learning (FL) has emerged as a decentralized approach that enables multiple nodes to collaboratively train mod-els without exchanging raw data, thereby protecting sensitive information. In this work, a Priva-cy-Preserving Federated Learning (PPFL) framework is introduced to provide secure, scalable, and efficient environmental data analysis across heterogeneous IoT systems. The proposed framework incorporates Differential Privacy (DP), Secure Aggregation, and Homomorphic En-cryption (HE) to ensure the protection of sensitive data during both communication and model updates. Experimental results show that the PPFL approach achieves a predictive accuracy of 97.6% with an RMSE of 0.203, surpassing recent FL-based techniques by up to 5.3%, while also maintaining strong privacy safeguards and reducing communication costs. This research presents a novel integration of differential privacy, homomorphic encryption, and edge computing within a unified federated learning framework for real-time environmental monitoring, effectively balanc-ing accuracy, privacy, and scalability.
Electric V2V Energy Transfer Using Bidirectional On-Board Converters Using Fuzzy Logic Controller
Authors: Mounika Baddula, Dr. P. Kowstubha
Abstract: Electric Vehicle-to-Vehicle (V2V) energy transfer is an emerging solution for optimizing charg-ing efficiency and extending the operational range of electric vehicles (EVs). This paper explores an intelligent energy-sharing framework utilizing fuzzy logic-based control for onboard convert-ers to enable seamless and adaptive power exchange between EVs. The proposed system dynam-ically adjusts energy flow based on real-time vehicle parameters such as battery state-of-charge, load demand, and grid availability. By leveraging fuzzy logic optimization, the system enhances energy transfer efficiency, minimizes power losses, and ensures a stable and reliable exchange process. Simulation results demonstrate the effectiveness of the proposed approach in improving energy utilization while maintaining vehicle battery health. This study contributes to the advance-ment of smart energy management in EV networks, paving the way for more sustainable and decentralized charging solution. The complete system is modelled and simulated using MATLAB/Simulink under various operating modes, including forward boost mode and reverse buck mode. Simulation results such as battery SOC, output voltage, output current, and power transfer characteristics are analyzed to evaluate system performance. The results demonstrate that the proposed V2V energy transfer system achieves efficient bidirectional power flow, reduced overshoot, improved voltage stability, and enhanced charging performance. The proposed ap-proach offers a reliable and efficient solution for future smart electric vehicle energy management systems and sustainable transportation.
Advanced Voice-Driven AI Frameworks For En-terprise Incident Analysis And Escalation Workflows
Authors: William Turner, Charlotte Evans, Amelia Scott, Grace Phillips, Henry Collins, Jeji Krishnan
Abstract: The increasing complexity of enterprise infrastructures, cloud computing environments, and distributed operational systems has created major challenges in incident management, escalation handling, and real-time operational support. Traditional support engineering methods often de-pend on manual monitoring and reactive troubleshooting approaches, which can result in delayed incident resolution, operational downtime, and reduced service reliability. Recent advancements in Artificial Intelligence (AI), Natural Language Processing (NLP), speech recognition, machine learning, and generative AI technologies have enabled the development of advanced voice-driven AI frameworks that support intelligent enterprise operations and automated escalation workflows. This research paper explores advanced voice-driven AI frameworks for enterprise incident analy-sis and escalation workflow management by integrating voice-enabled AI assistants, predictive analytics, intelligent automation, and machine learning models to improve operational monitoring, incident classification, troubleshooting assistance, and escalation coordination. The study empha-sizes the importance of Human-in-the-Loop (HITL) methodologies in which human experts supervise AI-generated recommendations and validate critical escalation decisions to ensure relia-bility, transparency, accountability, and operational accuracy. Voice-driven AI assistants allow support engineers and operations teams to interact with enterprise systems through natural lan-guage commands, thereby improving accessibility, reducing response time, and enhancing collab-orative decision-making during critical operational incidents. The paper further examines the applications of voice-driven AI systems in enterprise operations centers, cloud infrastructure management, cybersecurity incident response, and intelligent IT service management. The pro-posed framework provides several benefits including faster incident resolution, proactive anomaly detection, improved operational efficiency, optimized resource utilization, reduced downtime, and enhanced customer satisfaction. Additionally, the research discusses key challenges such as data privacy concerns, speech recognition accuracy, AI explainability, integration complexity, and cybersecurity risks associated with intelligent operational systems. Finally, the study highlights future research directions in adaptive AI systems, explainable voice interfaces, generative AI support agents, and autonomous enterprise operations, demonstrating how advanced voice-driven AI technologies can transform enterprise incident analysis and escalation workflows through the effective combination of intelligent automation and expert human oversight.
Enterprise Visual Diagnostics Using AI Image Processing And Intelligent Dashboard Analytics
Authors: Amelia Scott, Benjamin Lewis, Charlotte Evans, William Turner, Olivia Harris, Jeji Krishnan
Abstract: Enterprise environments are becoming increasingly complex due to the rapid growth of cloud computing, distributed infrastructure, and real-time operational systems, creating a strong need for intelligent monitoring and diagnostic solutions capable of analyzing visual operational data effi-ciently. This research presents an advanced framework for Enterprise Visual Diagnostics using Artificial Intelligence (AI) image processing and intelligent dashboard analytics to improve infra-structure monitoring, anomaly detection, and enterprise decision-making processes. The proposed system integrates computer vision techniques, deep learning models, optical character recognition (OCR), and intelligent analytics to automatically interpret dashboard screenshots, graphical moni-toring panels, visual alerts, and infrastructure performance indicators in real time. By utilizing convolutional neural networks (CNNs) and multimodal AI-based analytical methods, the frame-work identifies abnormal system behaviors, performance degradation patterns, security threats, and operational bottlenecks with minimal human intervention. The study further explores auto-mated incident classification, predictive maintenance support, and intelligent escalation workflows to enhance operational efficiency and reduce downtime in enterprise ecosystems. Experimental analysis indicates that AI-driven visual diagnostics significantly improve the speed and accuracy of incident detection while providing proactive operational insights compared to conventional monitoring approaches. Additionally, the research discusses important implementation considera-tions including scalability, data quality, explainability of AI models, integration with enterprise platforms, and cybersecurity challenges. The findings demonstrate that intelligent dashboard analytics and AI-based image processing can transform enterprise support operations by enabling automated troubleshooting, real-time situational awareness, and data-driven decision-making across cloud, hybrid, and distributed enterprise infrastructures, thereby contributing to the devel-opment of next-generation autonomous enterprise monitoring and visual intelligence systems.
