SOS++: Finding Smart Behaviors Using Learning and Evolution

Abstract

We present SOS++, a bioinspired method combining evolution and learning, allowing the automatic design of the controller of autonomous agents, described as a finite-state machine. The application of this method to well-known problems, for example the follow-up of a trail or the resolution of a maze, led to the emergence of some behaviors we could qualify as intelligent. Moreover, it is possible to use the method in a hierarchical way in order to obtain complex behaviors starting from a set of basic actions. We have used an algorithm which is a variation of reinforcement learning with a reward adapted to the degree of uncertainty of the performed prediction.