Advantages and Limitations of using Successor Features for Transfer in Reinforcement Learning

Open in new window