Goto

Collaborating Authors

 real-world rl application


we study VRTDC in the online Markovian setting, which covers many real-world RL applications that have online

Neural Information Processing Systems

We thank the reviewers for providing valuable comments. Below are point-to-point responses to the important questions. Markovian setting is not that significant. A: We respectfully disagree with the reviewer. Q2: Keep problem's condition number in the complexity result.