Manning College Of Information And Computer Sciences @ UMass

Machine Learning, Neural Networks, Information Retrieval, Computer Network and Security, Advanced NLP

SRM University, AP

Computer Science: Software Engineering, Deep Learning, AI, Operating Systems, Database, Software Engineering, Mathematics, English

Research Intern @ Carnegie Mellon University

Remote, India

• Analyzed ECG waveforms and employed a neural network pipeline with Keras Sequential API to forecast patients’ cardiovascular conditions, achieving a 90% accuracy rate in prediction.
• Developed a centralized web-based platform with React.Js and Flask to deploy proposed models that automated to assist in clinical diagnosis and prognosis of nearly 2000 patients and reduced manual analysis time by 24%.
• Carried out survival analysis of post-thoracic transplantation patients who used the Intra-aortic balloon pump and Impella device using required statistical tests.

Web Developer Intern @ 360 Research Foundation

Remote, India

• Developed a user-focused website tailored to client specifications, utilizing Figma for design and Bootstrap for front-end implementation to ensure a modern and responsive user experience.
• Built an intuitive e-commerce portal incorporating a DOM-based shopping cart and leveraging the latest web frameworks to enhance user engagement and provide an immersive shopping experience.
• Integrated new features that boosted revenue by approximately 10% and contributed to expanding the organization’s market presence.

A Multilayer Framework for Data-Driven Student Modeling

Published in Springer: Link

• This research focuses on leveraging student affective states to provide personalized learning in a blended learning environment. By integrating clustering techniques and an ensemble of machine learning models, we developed the MLF-DSM framework to tailor learning experiences through dynamic student performance assessment. SHAP, an Explainable AI approach, was employed to enhance the transparency and interpretability of the model.Our findings demonstrate the potential of the MLF-DSM framework to improve personalized education by making data-driven decisions more transparent and accessible to educators.

Detecting Long Non-Coding RNAs Responsible for Cancer Development

Published in IEEE Xplore: Link

• With recent findings highlighting the importance of lncRNAs in cancer progression, this research aims to identify key cancer-related lncRNAs across five cancer types: Bladder, Colon, Cervical, Liver, and Head & Neck. Utilizing lncRNA expression profiles from 12,309 lncRNAs and 2,221 patients from TCGA, We utilized feature selection algorithms, including Variance Threshold, L1-based, and Tree-based methods for differential analysis. Our study identified 234 critical lncRNAs capable of distinguishing between these cancer types.