A hybrid search agent in pommerman

In this paper, we explore the possibility of search-based agents in games with resource-intensive forward models. We implemented a player agent in the Pommerman framework and put it against the baseline agent to measure its performance. We implemented a heuristic agent and improved it by enabling depth-limited tree search in specific gameplay moments. We also compared different node selection methods during depth-limited tree search. Our result shows that depth-limited tree search is still viable when presented with inefficient forward models and exploitation-driven selection method is the most efficient in this specific domain.

[1]  Amos J. Storkey,et al.  Training Deep Convolutional Neural Networks to Play Go , 2015, ICML.

[2]  Shih-Chieh Huang,et al.  Investigating the Limits of Monte-Carlo Tree Search Methods in Computer Go , 2013, Computers and Games.

[3]  Dennis J. N. J. Soemers,et al.  Tactical Planning Using MCTS in the Game of StarCraft , 2014 .

[4]  Roberto Bruni,et al.  Models of Computation , 2017, Texts in Theoretical Computer Science. An EATCS Series.

[5]  Taku Komura,et al.  A finite state machine based on topology coordinates for wrestling games , 2011, Comput. Animat. Virtual Worlds.

[6]  Christian P. Robert,et al.  Monte Carlo Methods , 2016 .

[7]  M. Veldhuis,et al.  Artificial Intelligence techniques used in First-Person Shooter and Real-Time Strategy games , 2011 .

[8]  Mark H. M. Winands,et al.  Sequential Halving for Partially Observable Games , 2015, CGW/GIGA@IJCAI.

[9]  Martin Müller,et al.  Fuego—An Open-Source Framework for Board Games and Go Engine Based on Monte Carlo Tree Search , 2010, IEEE Transactions on Computational Intelligence and AI in Games.

[10]  Simon M. Lucas,et al.  A Survey of Monte Carlo Tree Search Methods , 2012, IEEE Transactions on Computational Intelligence and AI in Games.

[11]  Vincent Vidal,et al.  A Lookahead Strategy for Heuristic Search Planning , 2004, ICAPS.

[12]  Steven te Brinke,et al.  Monte Carlo Tree Search , 2014 .

[13]  Esslli Site,et al.  Models of Computation , 2012 .

[14]  Michael Buro,et al.  Improving hearthstone AI by learning high-level rollout policies and bucketing chance node events , 2017, 2017 IEEE Conference on Computational Intelligence and Games (CIG).

[15]  Julian Togelius,et al.  Monte Mario: platforming with MCTS , 2014, GECCO.

[16]  Y. Björnsson,et al.  Game-Tree Properties and MCTS Performance , 2011 .

[17]  Rémi Bardenet,et al.  Monte Carlo Methods , 2013, Encyclopedia of Social Network Analysis and Mining. 2nd Ed..