Bootstrap Learning of alpha-beta-Evaluation Functions (1993)
BibTeX
@MISC{Heinz93bootstraplearning,
author = {Alois P. Heinz and Christoph Hense},
title = {Bootstrap Learning of alpha-beta-Evaluation Functions},
year = {1993}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
We propose ff-fi-evaluation functions that can be used in game-playing programs as a substitute for the traditional static evaluation functions without loss of functionality. The main advantage of an ff-fi-evaluation function is that it can be implemented with a much lower time complexity than the traditional counterpart and so provides a significant speedup for the evaluation of any game position which eventually results in better play. We describe an implementation of the ff-fi-evaluation function using a modification of the classical classification and regression trees and show that a typical call to this function involves the computation of only a small subset of all features that may be used to describe a game position. We show that an iterative bootstrap process can be used to learn ff-fi-evaluation functions efficiently and describe some of the experience we made with this new approach applied to a game called malawi. 1 Introduction Game playing programs especially those for tw...
Keyphrases
ff-fi-evaluation function alpha-beta-evaluation function bootstrap learning game position main advantage traditional counterpart game-playing program introduction game significant speedup regression tree small subset new approach iterative bootstrap process classical classification traditional static evaluation function typical call time complexity
