Chess Coaching with Maia-2

• Course: STA 561D: Probabilistic Machine Learning
• Methods: Maia-2, python-chess, Elo-adaptive coaching, probabilistic move modeling, interactive web interface

This project built an interactive chess tutoring system combining Maia-2 (a human-like move prediction model) with python-chess for board logic and a lightweight local web interface. The tutor adapts its recommendations and practice pool to the selected Elo level, supports FEN loading, move-by-move coaching, position evaluation, and a regret-and-retry mechanism for comparing candidate moves.

• Executive Summary
• FAQ
• Technical Appendix