Multi-Emitter MAP-Elites: Improving quality, diversity and convergence speed with heterogeneous sets of emitters

Quality-Diversity (QD) optimisation is a new family of learning algorithms that aims at generating collections of diverse and high-performing solutions. Among those algorithms, MAP-Elites is a simple yet powerful approach that has shown promising results in numerous applications. In this paper, we introduce a novel algorithm named Multi-Emitter MAP-Elites (ME-MAP-Elites) that improves the quality, diversity and convergence speed of MAP-Elites. It is based on the recently introduced concept of emitters, which are used to drive the algorithm's exploration according to predefined heuristics. ME-MAP-Elites leverages the diversity of a heterogeneous set of emitters, in which each emitter type is designed to improve differently the optimisation process. Moreover, a bandit algorithm is used to dynamically find the best emitter set depending on the current situation. We evaluate the performance of ME-MAP-Elites on six tasks, ranging from standard optimisation problems (in 100 dimensions) to complex locomotion tasks in robotics. Our comparisons against MAP-Elites and existing approaches using emitters show that ME-MAP-Elites is faster at providing collections of solutions that are significantly more diverse and higher performing. Moreover, in the rare cases where no fruitful synergy can be found between the different emitters, ME-MAP-Elites is equivalent to the best of the compared algorithms.