Saved in:
Bibliographic Details
Main Authors: Orseau, Laurent, Munos, Remi
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2405.04407
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911879015694336
author Orseau, Laurent
Munos, Remi
author_facet Orseau, Laurent
Munos, Remi
contents We improve the proofs of the lower bounds of Coquelin and Munos (2007) that demonstrate that UCT can have $\exp(\dots\exp(1)\dots)$ regret (with $Ω(D)$ exp terms) on the $D$-chain environment, and that a `polynomial' UCT variant has $\exp_2(\exp_2(D - O(\log D)))$ regret on the same environment -- the original proofs contain an oversight for rewards bounded in $[0, 1]$, which we fix in the present draft. We also adapt the proofs to AlphaGo's MCTS and its descendants (e.g., AlphaZero, Leela Zero) to also show $\exp_2(\exp_2(D - O(\log D)))$ regret.
format Preprint
id arxiv_https___arxiv_org_abs_2405_04407
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Super-Exponential Regret for UCT, AlphaGo and Variants
Orseau, Laurent
Munos, Remi
Machine Learning
Artificial Intelligence
We improve the proofs of the lower bounds of Coquelin and Munos (2007) that demonstrate that UCT can have $\exp(\dots\exp(1)\dots)$ regret (with $Ω(D)$ exp terms) on the $D$-chain environment, and that a `polynomial' UCT variant has $\exp_2(\exp_2(D - O(\log D)))$ regret on the same environment -- the original proofs contain an oversight for rewards bounded in $[0, 1]$, which we fix in the present draft. We also adapt the proofs to AlphaGo's MCTS and its descendants (e.g., AlphaZero, Leela Zero) to also show $\exp_2(\exp_2(D - O(\log D)))$ regret.
title Super-Exponential Regret for UCT, AlphaGo and Variants
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2405.04407