Alpha-Mini: Minichess Agent with Deep Reinforcement Learning
This work addresses game-playing AI for a simplified chess variant, but it is incremental as it applies existing methods to a new domain.
The researchers tackled the problem of training an agent for Gardner minichess using deep reinforcement learning, achieving a near-perfect win rate of 0.97 against a random agent.
We train an agent to compete in the game of Gardner minichess, a downsized variation of chess played on a 5x5 board. We motivated and applied a SOTA actor-critic method Proximal Policy Optimization with Generalized Advantage Estimation. Our initial task centered around training the agent against a random agent. Once we obtained reasonable performance, we then adopted a version of iterative policy improvement adopted by AlphaGo to pit the agent against increasingly stronger versions of itself, and evaluate the resulting performance gain. The final agent achieves a near (.97) perfect win rate against a random agent. We also explore the effects of pretraining the network using a collection of positions obtained via self-play.