Exploring Adaptive MCTS with TD Learning in miniXCOM

TL;DR

This paper introduces MCTS-TD, an adaptive Monte Carlo tree search algorithm enhanced with temporal difference learning, which improves game performance without pre-training, demonstrated on miniXCOM.

Contribution

The paper presents a novel adaptive MCTS algorithm using TD learning that operates online without pre-training, applied successfully to miniXCOM.

Findings

MCTS-TD outperforms standard MCTS in miniXCOM.

Adaptive MCTS with TD learning reduces training time.

Improved decision-making in turn-based tactical games.

Abstract

In recent years, Monte Carlo tree search (MCTS) has achieved widespread adoption within the game community. Its use in conjunction with deep reinforcement learning has produced success stories in many applications. While these approaches have been implemented in various games, from simple board games to more complicated video games such as StarCraft, the use of deep neural networks requires a substantial training period. In this work, we explore on-line adaptivity in MCTS without requiring pre-training. We present MCTS-TD, an adaptive MCTS algorithm improved with temporal difference learning. We demonstrate our new approach on the game miniXCOM, a simplified version of XCOM, a popular commercial franchise consisting of several turn-based tactical games, and show how adaptivity in MCTS-TD allows for improved performances against opponents.