Benchmarking Supervised and Self-Supervised Mortality Risk Prediction Models on ECG Data
đź“‚GitHub Repository
A Hybrid Approach for Detecting AI-Generated Text in English
đź“‚GitHub Repository
Which Side Should We Scale? Revisiting Parameter Balance in Vision-Language Models
đź“‚GitHub Repository
Interpretable and Explainable Classification for Medical Data
This project explores interpretable and explainable machine learning techniques applied to both tabular and image-based data.
Time Series Modeling for Mortality Predictions
This project focuses on working with complex healthcare data from the Physionet 2012 Challenge dataset, which contains 48 hours of intensive care data used to predict patient mortality. It involves handling noisy, sparse, and irregularly-sampled multi-variate data through preprocessing, exploration, supervised learning, representation learning, and leveraging large language models (LLMs) for embedding extraction and analysis. Contributions