Progressive Strategies for Monte-Carlo Tree Search

By Guillaume M, J-B Chaslot, Mark H. M. Winands, H. Jaap Van Den Herik, Jos W. H. M. Uiterwijk, Bruno Bouzy. 2008

500 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Economics and Econometrics, Sociology and Political Science

View Paper: https://www.semanticscholar.org/paper/966b8c5b01f1eb4b7bb8f4a83ba3f1f1879f5250

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Economics and Econometrics

In Sociology and Political Science

Citations in the corpus

PACHI: State of the Art Open Source Go Program (2011, background): “(This is similar to the progressive bias [7], but not equivalent.”
COMPUTING “ELO RATINGS” OF MOVE PATTERNS IN THE GAME OF GO (2007, methodology): “Progressive widening was independently invented by Chaslot et al. (2007) Results indicate that using patterns in simulations and progressive widening both bring signiﬁcant improvements to the playing strength on 9 × 9.”
Single-Player Monte-Carlo Tree Search (2008, methodology): “In general, MCTS consists of four steps, repeated until time has run out [7].”
X-Armed Bandits (2010, unknown): “In Section 5 we compare the results of this paper with previous works.”
Parallel Monte-Carlo Tree Search (2008, methodology): “The three parallelization methods are implemented and tested in our Go program Mango [5] (mainly designed and constructed by Guillaume Chaslot), running on a multi-core machine containing 16 cores.”
Move Evaluation in Go Using Deep Convolutional Neural Networks (2014, methodology): “In addition, the MCTS engine utilised standard heuristics for computer Go: RAVE (Gelly & Silver, 2011), a UCT exploration strategy similar to Chaslot et al. (2008), and very simple rollouts based solely on 3 × 3 patterns (Huang et al., 2011).”
Computing Approximate Nash Equilibria and Robust Best-Responses Using Sampling (2011, unknown): “It is well-known that a perfectly rational Nash-Equilibrium strategy (NES) is not necessarily a bestresponse strategy (i.e., best counter-strategy) against strategies other than the rational strategy itself (Osborne & Rubinstein, 1994).”
Evaluating and modelling Hanabi-playing agents (2017, background): “1: The four steps of Monte Carlo Tree Search [10]”
Monte-Carlo Tree Search Solver (2008, methodology): “Therefore, we have chosen a method called Secure child [9].”
MoHex 2.0: A Pattern-Based MCTS Hex Player (2013, methodology): “as in Mango [10], we added a progressive bias term to the UCT formula.”
Multi-armed bandits with episode context (2011, methodology): “But computer Go has progressed further with trial-and-error development of more complex heuristics (e.g. (Chaslot et al. 2008)) for which there is little theoretical understanding of the type available for UCB1.”
Re-determinizing Information Set Monte Carlo Tree Search in Hanabi (2019, methodology): “• the use of heuristic rules to restrict the action space could naturally be extended with progressive un-pruning [30], using the learned evaluation function to decide which actions to un-prune.”
Growing Action Spaces (2019, background): “In the context of planning on-line with a model, there are a number of approaches that use progressive widening to consider increasing large action spaces over the course of search (Chaslot et al., 2008), including in planning for continuous action spaces (Couëtoux et al., 2011).”
Feature-Guided Black-Box Safety Testing of Deep Neural Networks (2017, methodology): “In this section, we compare our method with existing approaches, show convergence of the MCTS algorithm on limited runs, evaluate safety-critical networks trained on traffic light images, and counter-claim a recent statement regarding adversarial examples in physical domains.”