Pinned Loading

1. Hybrid Python/C++ sentiment classifier with a custom SGD training engine, automated model selection across 8 models, and sub-millisecond inference

   Python

2. Machine learning pipeline for heart disease probability estimation using Gradient Boosting, Random Forest, and Logistic Regression achieving AUC 0.954 across 13 clinical features.

   Python

3. Classifying Amazon product reviews by positive or negative sentiment for the SP26 INLS 642 Kaggle competition

   Python

4. Modular R pipeline for binary smoking status classification from clinical biomarkers, featuring XGBoost, stacked ensembles, and 10-fold CV tuning across 9 model families on 15,000 patient records.

   R

5. A containerized FastAPI backend that leverages language models to automatically classify medical specialties and extract clinical entities from unstructured transcription data.

   Python

6. A fast, stateless, and privacy-focused web application that calculates exact medication refill dates and tracks your remaining pill supply.

   HTML