Deep RL framed as continuous-time stochastic process

Arxiv CS.LG (machine learning)

The paper models deep RL as a continuous-time stochastic process, drawing on stochastic control theory. It provides a theoretical framework for analyzing RL dynamics in continuous environments.

reinforcement-learningdeep-learning

Arxiv CS.LG (machine learning)·Saket TiwariTejas KotwalGeorge Konidaris

7 days ago