Hemalatha D

@veltech.edu.in

Assistant Professor, Dept of CSE
Vel Tech Rangarajan Dr. Sagunthala R & D Institute of Science and Technology

Hemalatha D


               Download Compact PDF  
https://researchid.co/hema24294

RESEARCH INTERESTS

Deep Learning, Data Mining, Mobile Application Development
38

Scopus Publications

Scopus Publications

  • HELM-BRCA: Hybrid Embedding and Learning Model for BRCA Methylation Classification
    Hemalatha D, N Gomathi
    International Journal of Advanced Computer Science and Applications, 2026
    Breast cancer remains a highly heterogeneous disease for which it demands advanced computational techniques that can reveal significant biological patterns in high-dimensional epigenomic data. DNA methylation profiles generated by the Illumina HumanMethylation450 platform yield rich, clinically relevant signals but introduce significant analytical challenges due to their high dimensionality, sparsity, and nonlinear structure. This work presents a novel memory-efficient hybrid learning architecture that combines Truncated Singular Value Decomposition (SVD), a deep Autoencoder, and a multi-model ensemble classifier for boosting subtype classification performance using TCGA-BRCA methylation data. In order to circumvent memory limits and prevent system crashes, a probe-subset extraction strategy combined with variance-based feature selection was employed to ensure fast and safe data loading from the Xena repository. While the autoencoder extracts compact nonlinear manifold representations, SVD captures the global linear variance structure. Further, the fused latent space is modelled by an ensemble including Random Forest, XGBoost, and a lightweight Keras neural classifier that allows the system to exploit different decision limits and achieve robust generalization. The experimental investigation across several architectures demonstrates high predictive performance with ROC-AUC scores exceeding 0.99 and accuracies higher than 0.96 for Basic CNN and MLP models. Furthermore, the proposed hybrid ensemble improves stability and precision by outperforming traditional baselines and confirming the complementary nature of spectral and deep feature extraction. This study is suitable for large-scale biomedical data analytics scenarios. In conclusion, this work provides an efficient hybrid machine learning framework for breast cancer methylation study by offering a strong platform for improved prognostic modelling and development of epigenetic biomarkers.
  • AI-driven business insights: Unleashing growth potential with advanced analytics and predictive modeling
    Hemalatha Dakshinamurthi, Venkata Bharath Chakravarthi Kudumula, Ravi Surya Anantha Rishwanth Patthipati, Sai Mani Krishna Chilukuri, Pavan Kalyan Chennuru, Alex David
    Aip Conference Proceedings, 2026
  • Sehath connect revolutionizing e-healthcare with artificial intelligence and machine learning
    Hemalatha Dakshinamurthi, Mir Muzammil Ali, Nenishkanth Baskaran, Alex David, Ravikumar, Almas Begum
    Aip Conference Proceedings, 2026
  • Exploring the efficacy of SVM, KNN, and random forest in diabetes risk prediction
    Almas Begum, Alex David Swaminathan, Hemalatha Dakshinamurthi, Ruth Naveena Nadarajan, Rajathi Krishnamoorthi, Vijayalakshmi Veeranarayanan, S. Saranya
    Aip Conference Proceedings, 2026
  • HCRHE-NET: HIGH-CONFIDENCE RESIDUAL HYBRID ENSEMBLE NETWORK FOR BREAST CANCER DETECTION FROM TCGA-BRCA DNA METHYLATION DATA
    N. GOMATHI HEMALATHA D
    Journal of Theoretical and Applied Information Technology, 2026
    Due to noise, redundancy and uncertainty with respect to predictive confidence, proper classification of high-dimensional biological data remains a significant challenge. To overcome these shortcomings, this paper proposes a High-Confidence Residual Hybrid Ensemble (HCRHE) Network, which is a composite of residual learning, deep neural modelling, and confidence-conscious decision fusion to classify diseases with high confidence. The Cancer Genome Atlas (TCGA) contains large-scale data on DNA methylation that are susceptible to overfitting and unstable forecasts by using conventional deep learning models to evaluate the proposed methodology. To learn latent patterns due to reconstruction, HC-RHE architecture consists of a primary-prediction base multilayer perceptron (MLP), together with a residual learning path that is constructed using an autoencoder and residual MLP. A new machine called a confidence-based fusion technique allows the dynamical weighting of the base and residual prediction in terms of model certainty and makes adaptive decision-making. Also, forecasts with high confidence margins are retained and a high-confidence filtering process used, which maximizes reliability with minimal coverage loss. An accuracy-optimized threshold selection strategy is also provided in order to enhance the performance of classification further. Vast comparative experiments are conducted with the state-of-the-art deep learning baselines, including CNN, autoencoder-based classifiers, Dense DropConnect, residual CNN frameworks, and Basic MLP (Adam and SGD). The proposed HC-RHE has a much better accuracy at 98.7 compared to all other methods. These results prove success of confidence-aware residual fusion as they reflect continuous improvement over the best baseline CNN model. The proposed framework is, all in all, exceptionally promising to the field of clinical decision-support systems and gives a credible, intuitive, and high-confidence classification paradigm of high-dimensional biomedical data.
  • A transformer-based contrastive learning framework for early breast cancer detection using DNA methylation profiles
    Hemalatha D, Gomathi N
    Journal of the Chinese Institute of Engineers Transactions of the Chinese Institute of Engineers Series A, 2026
    Early detection of breast cancer through epigenomic signatures offers a promising path to improved prognosis. In this work, we propose EpiFormerNet, a novel deep learning model that leverages DNA methylation patterns in breast tissue to identify cancer and non-cancer samples with high accuracy. The model integrates pretraining with contrastive learning-based with a Transformer encoder to capture high-dimensional interactions between CpG sites and long-range dependencies. Attention mechanisms in the Transformer architecture enable the identification of biologically relevant methylation sites, enhancing interpretability and diagnostic usefulness. Training and validation on methylation data for 114 breast cancer patients and 23 normal controls are conducted with excellent classification performance and minimal overfitting. Through the combination of self-supervised learning, attention-based fusion of features and deep contextual modeling, EpiFormerNet exhibits stronger generalization performance in low-sample genomic diagnostics. The proposed work gives an accuracy of 96.2% and loss of 0.11 without overfitting and local minima.
  • Automated Model Optimization and Explainability for Early Detection of Breast Cancer Using Optuna and SHAP
    Hemalatha D, N Gomathi
    Proceedings Icses 2026 5th International Conference on Innovative Computing Intelligent Communication and Smart Electrical Systems, 2026
    Breast cancer is a multifactorial disease for women affected by genetic, hormonal, and environmental elements. The advanced genomic and molecular studies play a crucial role in increasing survival rate for early detection through different screening methods such as mammography, biopsy and ultrasound. This study utilises Machine Learning (ML) techniques that predict breast cancer with The Cancer Genome Atlas (TCGA) dataset. The XGBoost, LightGBM, Random Forest, and CatBoost were implemented to improve early prediction of breast cancer to improve the performance using ensemble methods and optimization algorithm. It further performs the dimensionality reduction using PCA and applies SMOTE in order to resolve the class imbalance problem. The Bayesian optimization framework was used to optimize the Optuna to improve the performance and also increase the interpretability of the model with information on significant genetic and clinical characteristics (SHapley Additive explanations). The critiqued outcome indicates that the performance of ensemble-based machine learning models especially XGBoost is effective in the early prediction and detection of breast cancer. Future research will focus on multi-omics data integrating with deep learning architectures and testing the model in real-time clinical data to enhance precision.
  • Browser-Based Real-Time Vehicle Tracking with Simulated V2I Communication Using Firebase Firestore
    Hemalatha D, B. Nithin, V. Yashvanth Goud
    Proceedings Icses 2026 5th International Conference on Innovative Computing Intelligent Communication and Smart Electrical Systems, 2026
    Real-time monitoring of the movements of vehicles and the status of infrastructure is a key component of the development of smart cities and intelligent transportation systems. This paper proposes a lightweight, scalable web-based vehicle tracking system with live vehicles in addition to simulated Vehicle-to-Infrastructure (V2I) communication. The system architecture consists of three web clients: a driver interface which broadcasts GPS coordinates via the Geolocation API of the browser, a user-friendly map interface written in Leaflet.js that shows the location of the vehicle in realtime, and an administrative control panel which can be used to simulate smart city infrastructure information, such as the status and countdowns of traffic lights. The synchronization of communication and data between these parts is organized by Google Firebase Firestore, a cloud-based NoSQL database that implements the low-latency real-time updates. This serverless technology reduces the complexity and cost of the backend and guarantees a high-level of scalability. The system manages a proof-of-concept of V2I applications through allowing the user in the map to see not only the moving vehicle but also contextual alerts (e.g., “Driver is waiting at a red light”) that are being pushed through the administrative panel. The work is a guiding template on building more sophisticated real-time logistics, ride-sharing, and smart traffic control programs with modern and available web technologies.
  • Smart EV Charging Locator and Queue Management
    Hemalatha D, Aakash K, Vinitha K, Visumathi J
    Proceedings Icses 2026 5th International Conference on Innovative Computing Intelligent Communication and Smart Electrical Systems, 2026
    The fluctuating lineups and wait times at charging stations are one of the key issues facing exponential growth in electric vehicles. They seriously interfere with the user's convenience. The present work proposes a web-based platform that enables users to pre-book specific cable slots, assists EV drivers in searching for nearby charging stations, and predicts wait times using a machine learning model. The system retrieves station and cable port information by querying OpenChargeMap and partner APIs, utilizes browser geolocation to determine the position of the user, and showcases alternatives on a Google Maps interface. The backend, which is built using Node.js and a MySQL database, is utilized along with a separate Python-based machine learning module that generates forecasts of waiting time, using XGBoost and Scikit-learn. For each station, users can see the estimated wait time and the journey time. Moreover, they can see and book the available cable time slots. We compare four strategies: (1) naive nearest-station (based only on distance), (2) journey-time aware, (3) queue-aware, and (4) booking + wait + travel. Simulations across multiple test zones show that the combined strategy decreases the average total time- travel plus waiting- by 23 % and increases cable use. The paper contributes by presenting a waittime prediction model, a comparison of decisionmaking techniques, mapping, machine learning, and slot booking. It further discusses possible future improvements, such as dynamic pricing and IoT-based real-time updates.
  • Detection of Artificial Profiles on Social Media Platforms Using Machine Learning and Natural Language Processing
    D. Hemalatha, Gaddam Rishitha, Sai Surya Siddhardha, Pothanaboyina Dimple Praharsha, Yedla Praveen
    Lecture Notes of the Institute for Computer Sciences Social Informatics and Telecommunications Engineering Lnicst, 2026
  • Machine Learning Models in Breast Cancer Prediction: Performance Evaluation and Insights
    Hemalatha D, N Gomathi
    Proceedings 3rd International Conference on Self Sustainable Artificial Intelligence Systems Icssas 2025, 2025
  • Evaluating the Performance of Fully Connected, Convolutional, and Recurrent Neural Networks in Diabetes Risk Prediction
    Almas Begum, Alex David S, Hemalatha D, Cyrilraj V
    2025 International Conference on Biomedical Engineering and Sustainable Healthcare Icbmesh 2025 Proceedings, 2025
  • Evaluating Logistic Regression and SVM for Enhanced Diabetic Retinopathy Screening
    Almas Begum, Alex David. S, Hemalatha. D, Rajathi. K, Ruth Naveena. N
    2025 International Conference on Intelligent Computing and Knowledge Extraction Icicke 2025, 2025
  • Federated Vision Transformer for Kidney Stone Diagnosis
    Almas Begum, Alex David S, Hemalatha D, Sivagami S, S. Benisha, Vijayalakshmi V
    Proceedings 3rd International Conference on Self Sustainable Artificial Intelligence Systems Icssas 2025, 2025
  • Gaming using different hand gestures using artificial neural network
    Prema S, G Deena, Hemalatha D, Aruna K B, Hashini S
    Eai Endorsed Transactions on Internet of Things, 2024
  • Performance Comparison of SVM and Logistic Regression in Breast Cancer Diagnosis Using the WDBC Dataset
    Hemalatha D, N Gomathi, Alex David S, Almas Begum, Ruth Naveena N.
    2024 Global Conference on Communications and Information Technologies Gccit 2024, 2024
  • A Comprehensive Assessment of Machine Learning Techniques for Temperature and Rainfall Forecasting: SVR vs. Decision
    Rajathi K, Alex David S, Ruth Naveena N, Ravi Kumar S, Almas Begum, Hemalatha D
    2024 1st International Conference on Sustainability and Technological Advancements in Engineering Domain Sustained 2024, 2024
  • Deep Learning Advances in Brain Tumor Classification: Leveraging VGG16 and MobileNetV2 for Accurate MRI Diagnostics
    Alex David S, Almas Begum, S Ismail Kalilulah, Ruth Naveena N, Rajathi K, Hemalatha D
    4th International Conference on Power Energy Control and Transmission Systems Harnessing Power and Energy for an Affordable Electrification of India Icpects 2024, 2024
  • Evaluating Machine Learning Models for Sustainability to Improve Electric Vehicle Range Prediction
    Alex David S, Ruth Naveena N, Almas Begum, Hemalatha D, Rajathi K., Vijayalakshmi V
    2024 1st International Conference on Sustainability and Technological Advancements in Engineering Domain Sustained 2024, 2024
  • Navigating Guest Sentiments: Comparative Analysis of Decision Tree and VADER for Hotel Review Analysis
    S Alex David, Almas Begum, N Ruth Naveena, D Hemalatha, S Ravi Kumar, V Vijayalakshmi
    2nd International Conference on Self Sustainable Artificial Intelligence Systems Icssas 2024 Proceedings, 2024
  • Deep Learning Techniques Advancements in Apple Leaf Disease Detection
    E Kannan, Carmel Mary Belinda M J, Alex David S, Ruth Naveena N, Almas Begum, Hemalatha D
    Procedia Computer Science, 2024
  • Predictive Management of EV Charging Stations Using Machine Learning
    Alex David S, Almas Begum, Hemalatha D, Senthil K, Monikaa R, Ruth Naveena N
    10th International Conference on Electrical Energy Systems Icees 2024, 2024
  • Skin Cancer Detection Using Deep Learning Technique
    Dilli Hemalatha, Kaile Nishi Latha, Pandala.Madhavi Latha
    2023 2nd International Conference for Innovation in Technology Inocon 2023, 2023
  • Iris Flower Species Detection using Machine Learning Technique
    Hemalatha D, Yaramala Tarunkumar Reddy, Lakshmikanth K, Shaik Chennuru Davusaheb Saajid
    2023 3rd International Conference on Intelligent Technologies Conit 2023, 2023
  • Early Detection of Breast Cancer with IoT: A Promising Solution
    Hemalatha D, Alex David S, Almas Begum, Potti Hemanth
    Proceedings of the 2nd IEEE International Conference on Advances in Computing Communication and Applied Informatics Accai 2023, 2023
  • Deep Learning-Based Lung Cancer Classification: Recent Developments and Future Prospects
    Almas Begum, S Alex David, D Hemalatha, Lavanya Sita Sai Kollipara
    Proceedings of the 2nd IEEE International Conference on Advances in Computing Communication and Applied Informatics Accai 2023, 2023
  • Identification of Fake and Spam Users on Social Networking Platforms
    G Rajesh, K Priya Kalaivani, D Hemalatha, V Prabhu, N Siva Rama Lingham
    2023 Intelligent Computing and Control for Engineering and Business Systems Iccebs 2023, 2023
  • Bitcoin Price Forecasting: A Comparison of LSTM and Feedforward Neural Network
    Alex David S, Almas Begum, Carmel Mary Belinda M J, Hemalatha D, Ruth Naveena N
    2023 IEEE International Conference on Research Methodologies in Knowledge Management Artificial Intelligence and Telecommunication Engineering Rmkmate 2023, 2023
  • Unlocking The Power Of Deep Learning: A Guide To Automatic License Number Recognition
    Hemalatha D, Almas Begum, S Alex David, Gaddam Rishitha, Pothanaboyina Dimple Praharsha, Chundi Nandini
    2023 1st International Conference on Optimization Techniques for Learning Icotl 2023 Proceedings, 2023
  • Impact of Scheduling Algorithms of Wireless Network Control System
    Almas Begum, D. Hemalatha
    Lecture Notes in Networks and Systems, 2022
  • E-seller Web Application for Vegetables using STS Framework
    Hemalatha D, Vengala Geetha Charan
    Proceedings of the 2022 International Conference on Innovative Computing Intelligent Communication and Smart Electrical Systems Icses 2022, 2022
  • Spotting Web Attack in Iot Networks Using Ensemble Deep Learning Technique
    Hemalatha D, Kanagala Sai Deepak, Edara Sravani, Muvida Dhananjaya
    Proceedings of the 2022 International Conference on Innovative Computing Intelligent Communication and Smart Electrical Systems Icses 2022, 2022
  • A Combined Deep CNN: LSTM with a Random Forest Approach for Breast Cancer Diagnosis
    Almas Begum, V. Dhilip Kumar, Junaid Asghar, D. Hemalatha, G. Arulkumaran
    Complexity, 2022
  • An ingenious fast-flux network building to minimize distributed denial of service assault in cloud computing
    D Hemalatha, Almas Begum
    Journal of Computational and Theoretical Nanoscience, 2018
  • Aneffective cloud parcelling in a load balancing model
    International Journal of Engineering and Technology Uae, 2018
  • Privacy study on images uploaded in social networks
    International Journal of Engineering and Technology Uae, 2018
  • Optimized adapt visualization network absurd iot congenital blockage innovator
    International Journal of Civil Engineering and Technology, 2017
  • Enhanced privacy settings in online content sharing websites using key distribution center (KDC)
    International Journal of Civil Engineering and Technology, 2017