RSPSA: Enhanced Parameter Optimisation in Games
- Levante Kocsis
- Csaba Szepesvari, Department of Computing Science; PI of AICML
- Mark H. M. Winands
Most game programs have a large number of parameters that are crucial for their performance. While tuning these parameters by hand is rather difficult, successful applications of automatic optimisation algorithms in game programs are known only for parameters that belong to certain components (e.g. evaluation-function parameters). The SPSA (Simultaneous Perturbation Stochastic Approximation) algorithm is an attractive choice for optimising any kind of parameters of a game program, both for its generality and its simplicity. It's disadvantage is that it can be very slow. In this article we propose several methods to speed up SPSA, in particular, the combination with RPROP, using common random numbers, antithetic variables and averaging. We test the resulting algorithm for tuning various types of parameters in two domains, poker and LOA. From the experimental study, we conclude that using SPSA is a viable approach for optimisation in game programs, especially if no good alternative exists for the types of parameters considered
Citation
L. Kocsis, C. Szepesvari, M. Winands. "RSPSA: Enhanced Parameter Optimisation in Games". Advances in Computer Games (ACG), January 2005.| Keywords: | machine learning |
| Category: |
BibTeX
@incollection{Kocsis+al:ACG05,
author = {Levante Kocsis and Csaba Szepesvari and Mark H. M. Winands},
title = {RSPSA: Enhanced Parameter Optimisation in Games},
booktitle = {Advances in Computer Games (ACG)},
year = 2005,
}Last Updated: March 07, 2007Submitted by Nelson Loyola
