k-Means Maximum Entropy Exploration

Alexander Nedergaard, Matthew Cook

Exploration in high-dimensional, continuous spaces with sparse rewards is an open problem in reinforcement learning. Artificial curiosity algorithms address this by creating reward…