Paper Details
This research paper explores the application of Logistic Regression to financial markets, specifically aiming to classify daily stock volatility based on historical price data.
Leveraging Python and scikit-learn, I built a machine learning pipeline that includes:
- Automated data fetching using
yfinance. - Feature engineering (calculating logarithmic returns and historical volatility).
- Data preprocessing using Z-score normalization.
- Hyperparameter tuning and regularization to optimize the model’s accuracy in predicting “high volatility” trading days.