JAJayanthi Arumugam

Jayanthi Arumugam

*This is an algorithmically generated author record based on publicly available information.

Jayanthi Arumugam has a total of 13 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    An optimized deep learning framework for brain tumor classification using magnetic resonance imaging

    by Komal Kumar Napa, Rajkumar Govindarajan, Senthil Murugan Janakiraman, Jayanthi Arumugam

    Accurate and interpretable classification of brain tumors in magnetic resonance imaging (MRI) scans plays a crucial role in early diagnosis and effective treatment planning. This study introduces a deep learning (DL) framework based on a customized YOLOv5m architecture integrated with a bidirectional feature pyramid network (BiFPN) for multi-class brain tumor classification. The integration of BiFPN enhances multi-scale feature fusion, improving detection across varied tumor types, while YOLOv5m ensures real-time inference capabilities. To mitigate class imbalance, a class-weighted cross-entropy loss is adopted. The model is evaluated on multiple performance metrics, achieving a test accuracy of 88.86%, precision of 88.70%, recall of 88.20%, and F1-score of 88.25%. It also reports a mean average precision (mAP@0.5) of 94.36%, with high class-wise average precision (AP) for glioma, meningioma, pituitary, and no-tumor categories. Computational time for training (12484.81 seconds) and testing (146.82 seconds) confirms the model’s feasibility for real-time clinical deployment. To support interpretability, gradient-weighted class activation mapping (Grad-CAM) is integrated for visualizing class-discriminative regions, helping clinicians understand the model’s predictions. A gradio-based user interface is also developed, enabling intuitive interaction with the system.

  • Open Access

    Comparative analysis of wind speed prediction: enhancing accuracy using PCA and linear regression vs. GPR, SVR, and RNN

    by Somasundaram Deepa, Jayanthi Arumugam, Raguraman Purushothaman, D. Nageswari, L. Rajasekhara Babu

    For power systems with significant wind power integration to operate in an efficient and dependable manner, wind speed prediction accuracy is crucial. Factors such as temperature, humidity, air pressure, and wind intensity heavily influence wind speed, adding complexity to the prediction process. This paper introduces a method for wind speed forecasting that utilizes principal component analysis (PCA) to reduce dimensionality and linear regression for the prediction model. PCA is employed to identify key features from the extensive meteorological data, which are subsequently used as inputs for the Linear Regression model to estimate wind speed. The proposed approach is tested using publicly available meteorological data, focusing on variables such as temperature, air pressure, and humidity. Popular models like recurrent neural networks (RNN), support vector regression (SVR), and Gaussian process regression (GPR) are used to compare its performance. Evaluation metrics such as root mean square error (RMSE) and R² are used to measure effectiveness. Results show that the PCA combined with Linear Regression model yields more accurate predictions, with an RMSE of 94.11 and R² of 0.9755, surpassing the GPR, SVR, and RNN models.


  • Open Access

    Evaluation of sequential feature selection in improving the K-nearest neighbor classifier for diabetes prediction

    by Rajkumar Govindarajan, Vidhyashree Balaji, Jayanthi Arumugam, Tsehay Admassu Assegie, Radha Mothukuri

    The K-nearest neighbor (KNN) classifier employs distance metrics to measure the distance between the test instance and the samples used in training. With smaller samples, the KNN classifier achieves higher accuracy with low computational time. However, computing the distance between the test instance and all training samples to determine the class of the test instance requires higher computational time for a high-dimensional dataset. This research employs sequential feature selection (SFS) to select the optimal feature for diabetes prediction while reducing the computational time complexity of the KNN classifier. The KNN classifier showed effectiveness with an accuracy rate of 84.41% with nine features. The performance of the KNN improves by 2.6% when trained on the optimal features selected with the SFS. The result revealed glucose level, blood pressure (BP), skin thickness (ST), diabetes pedigree function (DPF), age, and body mass index (BMI) as the most representative features in diabetes prediction. The KNN classifier gives higher accuracy with these features. However, insulin and the number of times a woman is pregnant do not show a significant effect on the KNN classifier.


  • Open Access

    Prediction of patient survival from heart failure using a cox-based model

    by Tsehay Admassu Assegie, Thulasi Karpagam, Sathya Subramanian, Senthil Murugan Janakiraman, Jayanthi Arumugam, Dawed Omer Ahmed

    The existing heart failure risk prediction models are developed based on machine learning predictors. The objective of this study is to identify the key risk factors that affect the survival time of heart patients and to develop a heart failure survival prediction model using the identified risk factors. A cox proportional hazard regression method is applied to generate the proposed heart failure survival model. We used the dataset from the University of California Irvine (UCI) clinical heart failure data repository. To develop the model we have used multiple risk factors such as age, anemia, creatinine phosphokinase, diabetes history, ejection fraction, presence of high blood pressure, platelet count, serum creatinine, sex, and smoking history. Among the risk factors, high blood pressure is identified as one of the novel risk factors for heart failure. We have validated the performance of the model via statistical and empirical validation. The experimental result shows that the proposed model achieved good discrimination and calibration ability with a C-index (receiver operating characteristic (ROC) of being 0.74 and a log-likelihood ratio of 81.95 using 11 degrees of freedom on the validation dataset.