ME-MCTS: Online Generalization by Combining Multiple Value Estimators

Authors: Hendrik Baier, Michael Kaisers

IJCAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments with three abstractors in four board games show significant improvements of ME-MCTS over MCTS using only a single abstractor, both for MCTS with random rollouts as well as for MCTS with static evaluation functions.
Researcher Affiliation Academia Hendrik Baier , Michael Kaisers Centrum Wiskunde & Informatica, Amsterdam {hendrik.baier, michael.kaisers}@cwi.nl
Pseudocode No The paper describes methods using equations but does not contain structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide concrete access to source code for the methodology described.
Open Datasets No The paper tests ME-MCTS in four different domains: Breakthrough, Knightthrough, Othello, and Rolit. However, it does not provide concrete access information for a publicly available or open dataset for these games in a formal dataset sense (e.g., no link, DOI, or specific citation to a dataset repository).
Dataset Splits No The paper does not provide specific dataset split information needed to reproduce the data partitioning.
Hardware Specification No The paper does not provide specific hardware details used for running its experiments.
Software Dependencies No The paper does not provide specific ancillary software details with version numbers needed to replicate the experiment.
Experiment Setup Yes All experiments allowed for 250 ms per move, in order to fairly expose the tradeoff between improving search with multiple estimators and the additional computational overhead of computing and combining abstractors. The implemented abstractors each have two parameters: a bias parameter that is used to compute their influence on every move choice, and an exploration factor for the exploration-exploitation tradeoff of the AUER algorithm. Vanilla MCTS has one parameter: the exploration factor of UCB1.