Programmed Bayesian Neural Network for multi-modal posterior distributions with a numeric approach of marginalisationImplemented Bayesian inference framework by Langevin-gradient based Markov Chain Monte Carlo (MCMC) samplingAddressed Class imbalance problems by combination of SMOTifed-GAN variant of SMOTE and MCMC sampling methodsExperimented the proposed framework on 11 benchmark imbalanced datasets; tested the same with other versions of SMOTE

Designed Machine Learning based model to predict proteins undergoing LLPS through supervised regression analysisModeled the predictor with Pfeature for computing wide protein and peptide characteristics based on their amino acid sequenceResearch into Cellular Homeostasis & function has been simplifed - thanks to model of the predictor that has been by effective

Mentored first year students for the course Intro. to Electrical Engg.(ELL100) on a weekly basis

Innovated yearly operations for online semester, achieving 98%YoY increase in watch-time and 71%YoY increase in views. Coordinated virtual conceptualisation for Annual Freshers introduction video, garnering 1.4M impressions & 155K+ views