Skip to Main content Skip to Navigation
Conference papers

Scaling MAP-Elites to Deep Neuroevolution

Cédric Colas 1 Vashisht Madhavan 2 Joost Huizinga 2 Jeff Clune 3
1 Flowers - Flowing Epigenetic Robots and Systems
Inria Bordeaux - Sud-Ouest, U2IS - Unité d'Informatique et d'Ingénierie des Systèmes
Abstract : Quality-Diversity (QD) algorithms, and MAP-Elites (ME) in particular, have proven very useful for a broad range of applications including enabling real robots to recover quickly from joint damage, solving strongly deceptive maze tasks or evolving robot morphologies to discover new gaits. However, present implementations of ME and other QD algorithms seem to be limited to low-dimensional controllers with far fewer parameters than modern deep neural network models. In this paper, we propose to leverage the efficiency of Evolution Strategies (ES) to scale MAP-Elites to high-dimensional controllers parameterized by large neural networks. We design and evaluate a new hybrid algorithm called MAP-Elites with Evolution Strategies (ME-ES) for post-damage recovery in a difficult highdimensional control task where traditional ME fails. Additionally, we show that ME-ES performs efficient exploration, on par with state-of-the-art exploration algorithms in high-dimensional control tasks with strongly deceptive rewards.
Complete list of metadata
Contributor : Cédric Colas Connect in order to contact the contributor
Submitted on : Wednesday, January 6, 2021 - 3:11:28 PM
Last modification on : Tuesday, January 25, 2022 - 3:14:19 AM
Long-term archiving on: : Wednesday, April 7, 2021 - 7:57:27 PM


Files produced by the author(s)




Cédric Colas, Vashisht Madhavan, Joost Huizinga, Jeff Clune. Scaling MAP-Elites to Deep Neuroevolution. GECCO 2020 - Genetic and Evolutionary Computation Conference, Jul 2020, Cancun / Virtual, Mexico. ⟨10.1145/3377930.3390217⟩. ⟨hal-03099878⟩



Les métriques sont temporairement indisponibles