Robust Losses for Learning Value Functions
Sign Up to like & getrecommendations! Published in 2022 at "IEEE Transactions on Pattern Analysis and Machine Intelligence"
DOI: 10.1109/tpami.2022.3213503
Abstract: Most value function learning algorithms in reinforcement learning are based on the mean squared (projected) Bellman error. However, squared errors are known to be sensitive to outliers, both skewing the solution of the objective and… read more here.
Keywords: learning value; losses learning; bellman error; robust losses ... See more keywords