Detail výsledku
Creating Action Heuristics for General Game Playing Agents
SCHIFFEL, S.
Monte-Carlo Tree Search (MCTS) is the most popular search algorithm used in General Game Playing (GGP) nowadays mainly because of its ability to perform well in the absence of domain knowledge. Several approaches have been proposed to add heuristics to MCTS in order to guide the simulations. In GGP those approaches typically learn heuristics at runtime from the results of the simulations. Because of peculiarities of GGP, it is preferable that these heuristics evaluate actions rather than game positions. We propose an approach that generates heuristics that estimate the usefulness of actions by analyzing the game rules as opposed to the simulation results. We present results of experiments that show the potential of our approach.
Goal Condition, Heuristic Function, Game State, Game Tree, General Game
@inproceedings{BUT163390,
author="TRUTMAN, M. and SCHIFFEL, S.",
title="Creating Action Heuristics for General Game Playing Agents",
booktitle="Computer Games, CGW 2015",
year="2016",
series="Communications in Computer and Information Science",
journal="Communications in Computer and Information Science",
volume="614",
number="1",
pages="149--164",
publisher="Springer Verlag",
address="Berlín",
doi="10.1007/978-3-319-39402-2\{_}11",
issn="1865-0929",
url="https://link.springer.com/chapter/10.1007/978-3-319-39402-2_11"
}