Master’s Thesis
Market Predictive Strategy (S&P 500),
Developed a quantitative trading strategy using machine learning models (XGBoost + ElasticNet) with regime detection and dynamic position sizing. Implemented rolling feature engineering, XGBoost-based feature selection, and time-series cross-validation with embargo to avoid data leakage. Integrated risk management components such as market-condition filters, volatility-based sizing, and regime-aware scaling.
Key results: Sharpe 1.11, Adjusted Sharpe 1.14, Sortino 1.36.
Email Classifier
Machine Learning Course Project, Built a machine learning classifier for corporate email categorization.
Performed text preprocessing, feature extraction, and model training in Python.