Invariant causal prediction for block MDPs

Generalization across environments is critical to the successful application of reinforcement learning (RL) algorithms to real-world challenges. In this work we propose a method for learning state abstractions which generalize to novel observation distributions in the multi-environment RL setting. W...

Cur síos iomlán

Sonraí bibleagrafaíochta
Príomhchruthaitheoirí: Zhang, A, Lyle, C, Sodhani, S, Filos, A, Kwiatkowska, M, Pineau, J, Gal, Y, Precup, D
Formáid: Conference item
Teanga:English
Foilsithe / Cruthaithe: Proceedings of Machine Learning Research 2020

Míreanna comhchosúla