Discretizing Continuous Action Space with Unimodal Probability Distributions for On-Policy Reinforcement Learning

Open in new window