Benchmarking Quality-Diversity Algorithms on Neuroevolution for Reinforcement Learning. (arXiv:2211.02193v1 [cs.NE])

We present a Quality-Diversity benchmark suite for Deep Neuroevolution in
Reinforcement Learning domains for robot control. The suite includes the
definition of tasks, environments, behavioral descriptors, and fitness. We
specify different benchmarks based on the complexity of both the task and the
agent controlled by a deep neural network. The benchmark uses standard
Quality-Diversity metrics, including coverage, QD-score, maximum fitness, and
an archive profile metric to quantify the relation between coverage and
fitness. We also present how to quantify the …

algorithms arxiv benchmarking diversity neuroevolution quality reinforcement reinforcement learning