Interactive Evolutionary Computation for Real Robot from Different Viewpoints of Observation

Abstract

In this paper, we describe a fast learning method for a mobile robot which acquires autonomous behaviors from interaction between an operator and a robot. We develop a behavior learning method ICS (Interactive Classifier System) using interactive evolutionary computation regard for the teaching cost and a mobile robot is able to quickly learn rules so that an operator can directly teach a physical robot. Also the ICS is a novel evolutionary robotics approach using an adaptive classifier system. We classify teaching methods as internal observation (Learner View) and external one (Teacher View) from viewpoints of obsevation, and investigate relationship between observation methods and the results. We have two experiments based on our teaching methods on a real world.