Building Moon: A Statistical Research Project on Daily Market Data

Over the last months I have been developing "Moon", a personal research project combining Python, historical market datasets and machine learning techniques.

The purpose of the project is to investigate whether daily market direction can be studied using statistical methods and feature engineering.

Moon is not intended to provide trading signals, investment recommendations or financial advice.

Current areas of research include:

• Historical data collection
• Feature engineering (ATR, SMA, volatility metrics)
• Multi-asset datasets
• Statistical validation
• Long-term performance evaluation
• Machine learning experiments

This blog will serve as a development diary documenting technical challenges, lessons learned and future improvements.

All content is published exclusively for educational and research purposes.