At NIT Sikkim, Iām an Electronics and Communication Engineering undergraduate focused on applied machine learning. My Kaggle work has placed in the top 6% in competitions involving mass-spectrometry molecule identification and F1 pit-stop prediction.
For PulmoCare, I built an audio preprocessing and prediction pipeline for cough-based tuberculosis detection, using mel spectrograms, VGGish embeddings, and wavelet scalograms. I trained and evaluated ResNet and VGGish models with k-fold cross-validation.
