Memory-Enhanced Evolutionary Robotics: The Echo State Network Approach

Cédric Hartland (1) , Nicolas Bredeche (1, 2, 3) , Michèle Sebag (3)

1 LRI - Laboratoire de Recherche en Informatique
2 TANC - Algorithmic number theory for cryptology
3 TAO - Machine Learning and Optimisation

Interested in Evolutionary Robotics, this paper focuses on the acquisition and exploitation of memory skills. The targeted task is a well-studied benchmark problem, the Tolman maze, requiring in principle the robotic controller to feature some (limited) counting abilities. An elaborate experimental setting is used to enforce the controller generality and prevent opportunistic evolution from mimicking deliberative skills through smart reactive heuristics. The paper compares the prominent NEAT approach, achieving the non-parametric optimization of Neural Nets, with the evolutionary optimization of Echo State Networks, pertaining to the recent field of Reservoir Computing. While both search spaces offer a sufficient expressivity and enable the modelling of complex dynamic systems, the latter one is amenable to robust parametric, linear optimization with Covariance Matrix Adaptation-Evolution Strategies.

Mots clés: artificial neural networks; echo state networks; evolutionary robotics; memory; behavior-based robotics

Domaines: Intelligence artificielle [cs.AI]; Réseau de neurones [cs.NE]

Cédric Hartland, Nicolas Bredeche, Michèle Sebag. Memory-Enhanced Evolutionary Robotics: The Echo State Network Approach. Congress on Evolutionary Computation (CEC 2009), 2009, Trondheim, Norway. ⟨inria-00413238⟩