@selmarsmit

Extreme Value Based Adaptive Operator Selection

, , , and . PPSN, volume 5199 of Lecture Notes in Computer Science, Springer, (2008)

Abstract

Credit Assignment is a crucial ingredient for successful Adap- tive Operator Selection. Instead of the average fitness improvement of newborn offspring, this paper proposes to use some empirical order statis- tics of those improvements, arguing that rare but highly beneficial jumps matter as much or more than frequent but small improvements. An ex- treme value based Credit Assignment is thus proposed, rewarding each operator with the best fitness improvement observed in a sliding window for this operator. This mechanism, combined with existing Adaptive Op- erator Selection rules, is investigated in an EC-like setting. First results show that the proposed method allows both the Adaptive Pursuit and the Dynamic Multi-Armed Bandit selection rules to actually track the best operators along evolution.

Description

Selmar

Links and resources

Tags