ObjectWorld

Researchers developed a nonparametric Bayesian method for inverse reinforcement learning that handles diverse expert demonstrations by using a Dirichlet Process prior to infer latent reward types. The…