How ngnma/stock-return-prediction-using-machine-learning works
This is a notebook-based stock-return prediction project. Its documented workflow uses historical asset and S&P 500 data to create technical features, select features, compare and tune OLS, SVR, and Random Forest models, then evaluate predictions with time-aware validation and a chronological holdout. The README documents these stages, but no notebook source was provided, so internal call structure and how the listed result artifacts are produced are not established.
Main entry points: stock_return_prediction.ipynb Model selection
Components 14
Market data
- Historical market data
Feature and model development
- Technical features
notebooks/stock_return_prediction.ipynb - Feature selection
notebooks/stock_return_prediction.ipynb - OLS model
notebooks/stock_return_prediction.ipynb - SVR model
notebooks/stock_return_prediction.ipynb - Random Forest
notebooks/stock_return_prediction.ipynb - Hyperparameter tuning
notebooks/stock_return_prediction.ipynb
Model evaluation
- Model selection
notebooks/stock_return_prediction.ipynb - Time-series CV
notebooks/stock_return_prediction.ipynb - Mean baseline
notebooks/stock_return_prediction.ipynb - Final evaluation
notebooks/stock_return_prediction.ipynb
Other components
- Project user
- Model results
results/model_results.txt - Project report
reports/report.pdf
Connections 13
- Project user → Historical market data: provides context
README.md - Historical market data → Technical features: forms input
README.md - Technical features → Feature selection: informs selection
README.md - Feature selection → OLS model: selects features
README.md - Feature selection → SVR model: selects features
README.md - Feature selection → Random Forest: selects features
README.md - OLS model → Model selection: compared during
README.md - SVR model → Model selection: compared during
README.md - Random Forest → Model selection: compared during
README.md - Model selection → Hyperparameter tuning: uses tuning
README.md - Model selection → Time-series CV: uses validation
README.md - Model selection → Mean baseline: compares against
README.md - Model selection → Final evaluation: precedes evaluation
README.md