Comparing model performance across equities, rates, FX, and commodities. When do neural networks work? When do simple models outperform?
This repository contains two machine learning projects examining price forecasting across different asset classes. Each project isolates a specific question: Which models work best under different market regimes?
Central finding: Model choice depends entirely on data characteristics. Neural networks excel at smooth trends; tree-based methods survive regime shifts.
Youssef LOURAOUI
- Email: youssef.louraoui@essec.edu
- LinkedIn: linkedin.com/in/youssef-louraoui
- ResearchGate: researchgate.net/profile/Youssef-Louraoui-2
- GitHub: github.com/Gimkhana
MIT License. See individual project LICENSE files for details.