🎓 Final-year Financial Engineering student at HKUST | 📊 Quant & Fintech Enthusiast | 🤝 Community Leader
Welcome to my GitHub profile! I'm passionate about applying data-driven solutions to real-world financial problems. Whether it’s building robust statistical arbitrage strategies, developing explainable credit scoring models, or exploring crypto market dynamics, I enjoy combining mathematical rigour with hands-on coding.
- Languages: Python, C++, SQL, R, Pine Script, LaTeX
- Libraries: pandas, NumPy, scikit-learn, statsmodels, SciPy, yfinance, etc.
- Tools: Git, VSCode,TradingView, Overleaf
📊 Kalman Filter-Based Pairs Trading
Designed a dynamic statistical arbitrage strategy using Kalman Filters to model time-varying hedge ratios in cointegrated Consumer Staples stock pairs.
⚙ Python, pykalman, statsmodels, Yahoo Finance API
📈 Sharpe Ratio 1.26, Max Drawdown -3.5%, CAGR 6.7% (2022–2024 OOS)
🔍 Fraud Detection with Machine Learning
Built and evaluated models (Logistic Regression, Random Forest, XGBoost, Neural Network) on a highly imbalanced dataset (~0.8% fraud rate). Prioritised recall and AUC for rare event detection.
⚙ Python, scikit-learn, imbalanced-learn, XGBoost
📈 Neural Network achieved 97.5% recall and AUC of 0.9964
💡 Transparent Credit Scoring with Explainable ML
Developed an interpretable credit scoring model using SHAP to explain Random Forest predictions.
⚙ Python, scikit-learn, XGBoost, SHAP
📈 ~80% accuracy, Macro F1-score 0.79; generated actionable insights on key financial indicators
- Project Assistant @ Tech Nine Limited: Delivered QR code and smart access systems for residential and commercial use.
- Summer Intern @ OCBC (COO Office): Working on operational excellence initiatives in 2025.
- Student Leader @ HKUST: President of UG Hall VI's Connection Team, organising events for over 100+ residents.
"Data is the new oil, but only if refined with skill and purpose."
Thanks for visiting!