Angel – Chess Engine
One project that I am currently working on is building a chess engine called Angel. Inspired by my love for the game of chess and the battle of chess engines Stockfish vs AlphaZero. The goal is to build a chess engine that fully understands the rules of the game of chess and can evaluate positions using portable game notation (PGN). Then use backward induction, random forest regression, and machine learning to optimize the best possible move for the engine.
Latest Update:
All moves generated, including castling, en passant, and pawn promotion. All PGN is complete to make moves just need to add ‘+’ when the move puts the opposite king in check. Same thing with checkmate.
Next Update:
I found a bug that once you castling rooks on the same rank can’t move along that rank, looking to fix that bug. Next up, actually generating an AI to make a move, most likely have the CPU make a random selected valid chess move. Add intelligence later. Also looking to optimize looking for all possible moves each turn. (Python is a slower programming language, this can be an issue trying for the engine, but python is great for machine learning).
Check out my Progress on GitHub
Here you can find the code for my chess engine and even test it out!
Pingback: Degree of Randomness vs Skill - ryantvackner
Pingback: Your Decision Making is Determined by Your Time Horizon - ryantvackner