Popular repositories Loading

1. Bayesian_Final_Project_May_the_code_be_with_you Bayesian_Final_Project_May_the_code_be_with_you Public

   Team May the Code Be With You presents a Bayesian machine learning portfolio optimization project combining Black-Litterman, Hidden Markov Models, FinBERT-based NLP sentiment beliefs, and out-of-sa…

   Jupyter Notebook

2. Predicting-Deforestation-Risk-Economic-Environmental-and-Political-Drivers Predicting-Deforestation-Risk-Economic-Environmental-and-Political-Drivers Public

   Machine learning study of global deforestation drivers: clustering + classification showing economic development, not political ideology, predicts forest loss. Random Forest, 91.9% accuracy.

   HTML

3. Forecasting-Brent-Crude-Oil-Prices-Using-Time-Series-Models Forecasting-Brent-Crude-Oil-Prices-Using-Time-Series-Models Public

   A multi-model time series study forecasting Brent crude oil prices (2003–2025) using supply and demand drivers. VECM and tuned Prophet beat the naive benchmark by ~40%.

   Jupyter Notebook

4. Cash-Flow-Based-Mispricing-Signals-in-US-Equity-Returns Cash-Flow-Based-Mispricing-Signals-in-US-Equity-Returns Public

   Replicated empirical asset-pricing study on value-glamour and accruals mispricing (Desai et al., 2004); constructed CFO/P and accrual signals and tested their predictive power for future stock returns

   Jupyter Notebook

5. DRW-crypto-market-prediction DRW-crypto-market-prediction Public

   Top-100 of ~1,000 teams in the DRW Crypto Market Prediction Kaggle competition. Predicts short-term crypto price movements using feature clustering, purged time-series CV, and Ridge/XGBoost models.

   Jupyter Notebook