Developed ML-based models to predict return direction and benchmarked performance against traditional strategies using walk-forward testing to mitigate overfitting.

Focus: ML in finance, signal evaluation

Tools: Python, Scikit-learn, Feature Engineering

Key Outcomes:

• Designed leakage-safe ML pipelines for financial time series
• Applied walk-forward validation to test out-of-sample robustness
• Compared ML signals against baseline rule-based strategies

Links:

• Link