Not Logged In

Integrating Factorization Ranked Features in MCTS: an Experimental Study

Recently, Factorization Bradley-Terry (FBT) model is introduced for fast move prediction in the game of Go. It has been shown that FBT outperforms the state-of-the-art fast move prediction system of Latent Factor Ranking (LFR). In this paper, we investigate the problem of integrating feature knowledge learned by FBT model in Monte Carlo Tree Search. We use the open source Go program Fuego as the test platform. Experimental results show that the FBT knowledge is useful in improving the performance of Fuego.

Citation

C. Xiao, M. Müller. "Integrating Factorization Ranked Features in MCTS: an Experimental Study". Workshop on Computer Games (CGW), Springer, Cham, (ed: Cazenave T., Winands M., Edelkamp S., Schiffel S., Thielscher M., Togelius J.), 705, pp 34-43, July 2016.

Keywords:  
Category: In Workshop
Web Links: DOI
  Springer

BibTeX

@misc{Xiao+Müller:CGW16,
  author = {Chenjun Xiao and Martin Müller},
  title = {Integrating Factorization Ranked Features in MCTS: an Experimental
    Study},
  Booktitle = {Communications in Computer and Information Science},
  Publisher = {Springer, Cham},
  Editor = {Cazenave T., Winands M., Edelkamp S., Schiffel S., Thielscher M.,
    Togelius J.},
  Volume = "705",
  Pages = {34-43},
  booktitle = {Workshop on Computer Games (CGW)},
  year = 2016,
}

Last Updated: June 30, 2020
Submitted by Sabina P

University of Alberta Logo AICML Logo