Los Angeles, CA
March 31, 2009 to April 2, 2009
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/CSIE.2009.875
In this paper, we investigate an integration of best population pool and social learning, utilizing evolutionary neural networks. The experiments are divided into several intervals, and we keep the best player from each interval in the best population pool (BP-pool). Social learning allows poor performing players to learn from those players, which are playing at a higher level. The feed forward neural networks are evolved via evolutionary strategies and no knowledge incorporate in this work. The evolved neural network players had played against a rule-based player, Gondo, only at the beginning of the match. The remainder of the game then copied by another Gondo and they continued the game by playing against themselves. Learning is taking place in this experiment, where the slope of the regression line is significantly difference from zero.
Razali Yaakob, "An Integration of BP-Pool and Social Learning in the Opening of Go", CSIE, 2009, 2009 WRI World Congress on Computer Science and Information Engineering, CSIE, 2009 WRI World Congress on Computer Science and Information Engineering, CSIE 2009, pp. 636-640, doi:10.1109/CSIE.2009.875