MEENAKSHI M
@srmist.edu.in
Scopus Publications
- Intrusion Detection in Wireless Sensor Networks Using a Machine Learning-Driven Cyber Security Framework
M. Meenakshi, M. Mageshwari, S. P. M. Kannan, P. Veeramanikandan, K. Manivannan, Elangovan Muniyandy, R. Mohan Kumar
SN Computer Science, 2026 - AI-Driven Adaptive BCIs for Personalized Therapy
A. Rajalakshmi, G. SudhaAnanthi, Mohammad Sabeer, T. C. Manjunath, Yasmeen, M. Meenakshi
Brain Computer Interfaces for Neurorehabilitation, 2026
Artificial intelligence (AI)-driven adaptive brain-computer interfaces (BCIs) represent one of the most transformative frontiers at the intersection of neuroscience, computer science, and human–machine interaction. Traditional BCIs have long relied on a fixed mapping between brain signals and machine responses, but the challenge with these systems has been their limited adaptability to the natural variability of neural activity, user states, and contextual environments. Every individual's brain activity is highly dynamic, influenced by cognitive load, fatigue, emotions, attention, and even external environmental distractions. Fixed-model BCIs often struggle to deliver consistent performance across different users or even within the same user over time. Adaptive BCIs, enhanced with AI techniques such as machine learning (ML), deep learning, and reinforcement learning, overcome these challenges by continuously learning from real-time brain activity, adjusting algorithms on the fly, and personalizing the interaction model for each individual. - HAA: A Novel Hybrid Authentication Scheme for VANETs Integrating Certificate-Based and Identity-Based Cryptography with Advanced Attack Detection
N Gopinath, K. Lakshmi Narayanan, S. Sageengrana, M. Meenakshi
2025 International Conference on Computing and Communication Technologies Iccct 2025, 2025
In order to improve traffic efficiency and road safety, intelligent transportation systems rely on Vehicular Ad Hoc Networks (VANETs), which allow for encrypted communication. Nevertheless, strong authentication procedures are required since VANETs are dynamic and hence vulnerable to security risks including message manipulation, impersonation, and Sybil attacks. Current systems encounter issues like slowness, computational burden, and scalability constraints; these affect both identity-based and certificate-based authentication. Using a combination of certificate-based and identity-based cryptography, as well as an attack detection module based on machine learning, this study presents a new Hybrid Authentication Architecture (HAA). To guarantee security and efficiency, the HAA uses certificate-based mechanisms for important safety messages and lightweight identity-based procedures for everyday communications. With a latency reduction of 60%, computational overhead reduction of 25%, and support for up to 8,000 vehicles with an attack detection accuracy of 98%, the simulation results show that HAA is superior. With the possibility of incorporation into 5G-enabled systems and post-quantum cryptography frameworks in the future, the suggested architecture provides a scalable and efficient solution for secure VANET communications. - Enhancing urban mobility: machine learning-powered fusion approach for intelligent traffic congestion control in smart cities
Ankur Chaudhary, M. Meenakshi, Soma Sharma, Mahbubur Rahman, S. Srinivasan
International Journal of System Assurance Engineering and Management, 2025 - Enhanced Bitcoin Price Prediction using RNN-GRU Algorithm with Optimized Parameters: Overcoming LSTM Ambiguities for Improved Accuracy and Efficiency
M. Kathiravan, Meenakshi M, Kaviya M, S. Sreesubha, Sathyadurga. V, V. Neela Gandhi
3rd International Conference on Intelligent Data Communication Technologies and Internet of Things Idciot 2025, 2025
Bitcoin's decentralised character and introduction in January 2009 can be mostly attributed to Satoshi Nakamoto, the unidentified designer of the digital money. Unlike regular money, Bitcoin has no value in and of itself and is not supported by any government or financial institution. The “mining” process consumes a lot of resources thus it cannot manage trades without a network of computers, often known as “nodes” or “miners.” The Bitcoin blockchain, a catalogue of all the past transactions, is maintained by these nodes. Like equities, bitcoin is getting more and more well-known even if it is somewhat erratic as an investment. Its value varies greatly; hence it is difficult to predict when prices will adjust. Since Bitcoin is so erratic, automated approaches are growing more crucial for future projections regarding it. Often utilised for prediction, long short-term memory (LSTM) networks have constraints. LSTM focusses on short-term input and has a complex and occasionally ambiguous design, thus some claim it does not operate well when prior data is not significant to current forecasts. This has caused some to wonder if one can forecast Bitcoin's price using this approach. Made as a substitute for standard RNNs, the Gated Recurrent Unit (GRU) helps to prevent the fading gradient issue. Bitcoin price projections made by the GRU design are far more accurate than those derived from more antiquated methods. - Federated Learning and Blockchain-Based Hybrid Model for Stock Price Prediction and Financial Market Forecasting
M. Meenakshi, L. Vijayakumar, Meenu Baliyan, V. Venkata Ramana, B. Amarnath Reddy, Mini Srivastava
Proceedings of 10th International Conference on Communication and Electronics Systems Icces 2025, 2025 - A Framework Design of ML Classifier Algorithm for Retrieve the Information about the Drugs and its Quality
M Meenakshi, S. Ruban, T J Nandhini, S Loganayagi, Raad Muhammed Sayed, Enas Hassan, Imad Al-Din Hussein Al-Saidi
2024 4th International Conference on Advance Computing and Innovative Technologies in Engineering Icacite 2024, 2024
Every day, pharmaceutical firms have to do extensive product analyses. The same product is frequently registered more than once in several systems, each using unique properties. For these businesses, precise and high-quality information is crucial, especially considering the nature of the pharmaceuticals these goods contain. The hypothesis of this research project is that machine learning provides a workable way to effectively combine different data sources and match records pertaining to the same product. It is no longer feasible to match records by hand because of the extraordinarily large volume that has to be processed. In a large data context, this paper presents a framework for matching pharmacological records using machine learning approaches. The framework trains machine learning by utilising specified criteria for record matching. These models are then tested by forecasting matches between records that deviate from these preset guidelines. Lastly, by creating a huge variety of record combinations and forecasting matches based on them, the system mimics the production environment. The findings show that although the training datasets produce good results—the top model has an average accuracy of about $85 \\%$—the production environment has its own set of difficulties. In spite of this, matches that depart from established norms are effectively predicted by the framework. These results highlight how machine learning can be used to perform record matching tasks, as manual processing at this size is impossible. - Early Identification of as Disorder using Machine Learning Based Classifier System Implementation
G Karthika Priya Dharshini, M Meenakshi, T J Nandhini, Eppili Jaya, Riyad E. Abed, Fay Al-Taee, Abdul Hussein Mohsen
2024 4th International Conference on Advance Computing and Innovative Technologies in Engineering Icacite 2024, 2024
In this work, we utilized resting-state functional connectivity (FC) data to construct diagnostic classifiers using machine learning classifiers like random forest algorithms on four different autism spectral disorders (ASD) samples with a total of participants, ranging in age from 6 to 18 years. We purposefully altered the makeup of each sample to vary the degrees of gender diversity and symptom intensity, allowing for some overlap between groups and accounting for the wide range of ASD symptoms. Each sample had specific inclusion requirements, such as: all genders with an open-ended severity range; male participants with an open-ended severity; all genders with an emphasis on higher severity levels; and male participants only with higher severity levels. There were subjects in each subset (ASD, TD), carefully matched for head motion and age, of whom were assigned to training and to validation. Within every sample, random forest classifiers were trained. Sample through sample exhibited classification accuracies of corresponding to after testing in validation samples. Important factors affecting classifier performance were shown to vary among sample sets, including the connections within the cingulo-opercular task control (COTC) network and the interactions between COTC ROIs and the default mode and dorsal attention networks. Our findings demonstrate how difficult it can be to develop diagnostic classifiers based on the characteristics of ASD samples. The efficacy of classifiers is enhanced with greater uniformity about gender and symptom intensity. Nonetheless, it is crucial to recognize the inherent variability of ASDs, as relying exclusively on performance metrics may not be sufficient to accurately assess the usefulness of classifiers in practical applications. - Smart Agriculture: Soil Aggregate Stability Classification for Damaged Crops in India
M. Meenakshi, R. Naresh
International Journal on Advanced Science Engineering and Information Technology, 2023
Soil health is the most important element in a stable farm environment in soil-based agriculture. Soil aggregate stabilization is man-datory for soil characteristics influencing crop yield and stability. The study was conducted on Tamilnadu delta areas where the alluvial and black soil types for rabi and Kharif crops are used, and soil parameters are analyzed. This study aims to provide an overview of the mechanisms and aggregate-forming agents using ensemble methods. It is difficult to assess and analyze the aggregate stability. However, the most popular farming methods used in commercial crop yields, including artificial fertilizers and monocul-tures, can weaken the soil throughout the term, resulting in a sequence of issues that necessitate using many more man-made inputs, which contribute to global warming. The soil type's qualities and functions in predicting the crop type that can be grown under spe-cific soil conditions. Remote monitoring of soil parameters can change agricultural practices and boost productivity. We suggest a process in this article for classifying soil based on micro and macro-nutrients and predicting the form of the crop that can be grown in that type of soil. The results obtained were compared to the standardized maximum point for specific crops, and crop inputs var-ied depending on the variations. Several ensemble methods have been used, such as the bagging meta-estimator, Ada Boost, and XGB. On the held-out dataset, the bagging models estimated an accuracy of 98 percent, showing the technological viability of differ-ent soil types. - Automatic Emoji Generation using Inception V3
Ambati Himasree, Kada Sushil Kumar, M. Meenakshi
Proceedings of the 2nd International Conference on Edge Computing and Applications Icecaa 2023, 2023
Internet and communication technologies are evolving quickly today. As a result, communication is simpler than it always was. This study examines the usage of emojis for real-time emotion recognition. The findings indicate that while there are disparities in the types of emojis used, there are no differences in the frequency of use between genders. Emojis were mentioned as a way for people of both genders to give their communications more depth. Since they make both genders feel at ease when utilizing online communication tools, emojis are also not regarded to be gender specific. The current study focuses on how emojis, which display images of facial expressions, assist people in recognizing emotions. Additionally, it develops the measurement parameters. - Deep Learning Techniques for Spamming and Cyberbullying Detection
M Meenakshi, P Shyam Babu, V Hemamalini
Proceedings of the 1st IEEE International Conference on Networking and Communications 2023 Icnwc 2023, 2023 - A Cloud based Improved File Handling and Duplicate Removal using MD5
M. Kathiravan, R. Logeshwari, S. Pavithra, M. Meenakshi, V. Sathya Durga, M. Vijayakumar
Proceedings of the 3rd International Conference on Artificial Intelligence and Smart Energy Icais 2023, 2023 - Detecting Phishing Websites using Machine Learning Algorithm
M. Kathiravan, V. Rajasekar, Shaik Javed Parvez, V. Sathya Durga, M. Meenakshi, S. Gowsalya
Proceedings 7th International Conference on Computing Methodologies and Communication Iccmc 2023, 2023 - Soil health analysis and fertilizer prediction for crop image identification by Inception-V3 and random forest
M. Meenakshi, R. Naresh
Remote Sensing Applications Society and Environment, 2022 - Review on Health and Productivity Analysis in Soil Moisture Parameters
M. Meenakshi, R. Naresh
Lecture Notes in Networks and Systems, 2022 - Prediction of soil moisture root zone health in Artificial Neural Network
M. Meenakshi, R. Naresh
4th International Conference on Recent Trends in Computer Science and Technology Icrtcst 2021 Proceedings, 2022 - Efficient study of smart garbage collection for ecofriendly environment
Journal of Green Engineering, 2020 - Smart home: Security and acuteness in automation of IOT sensors
M.Meenakshi, R.Naresh, S.Pradeep
International Journal of Innovative Technology and Exploring Engineering, 2019
