The ChessAIThon project (2025-1-ES01-KA220-VET-000354329) is co-funded by the European Union. The views and opinions expressed in this publication are those of the author(s) only and do not necessarily reflect those of the European Union or the Spanish Service for the Internationalisation of Education (SEPIE). Neither the European Union nor the National Agency SEPIE can be held responsible for them.
Table of Contents
Deep Learning Integration (Neural MCTS)
The CNN acts as the policy network , predicting a probability vector ($p$) over promising moves. Instead of MCTS running random simulations, it uses these probabilities to focus on the 3 to 10 most promising moves in a position, allowing for a much deeper game simulation in high-quality variations. This integration allows the ChessMarro model to achieve the necessary strategic depth in the ChessAIthon competition.