A Potential of Evolutionary Rule-based Machine Learning for Real World Applications
Keiki Takadama · 2015
This paper explores a potential of Evolutionary Rule-based Machine Learning (ERML) by showing how ERML succeeds in real world applications. Generally, ERML is defined as a method that integrates (rule-based) machine learning with evolutionary computation, where the former method contributes to a local search while the latter method contributes to a global search. From such an integrated feature of ERML, one of the fundamental interests in ERML is how to control interactions between learning and evolution to produce a performance that cannot be achieved by either of these methods alone.