Abstract
We consider the off-policy evaluation problem of reinforcement learning using deep convolutional neural networks. We analyze the deep fitted Q-evaluation method for estimating the expected cumulative reward of a target policy, when the data are generated from an unknown behavior policy. We show that, by choosing network size appropriately, one can leverage any low-dimensional manifold structure in the Markov decision process and obtain a sample-efficient estimator without suffering from the curse of high data ambient dimensionality. Specifically, we establish a sharp error bound for fitted Q-evaluation, which depends on the intrinsic dimension of the state-action space, the smoothness of Bellman operator, and a function class-restricted χ2-divergence. It is noteworthy that the restricted χ2-divergence measures the behavior and target policies' mismatch in the function space, which can be small even if the two policies are not close to each other in their tabular forms. We also develop a novel approximation result for convolutional neural networks in Q-function estimation. Numerical experiments are provided to support our theoretical analysis.
Original language | English (US) |
---|---|
State | Published - 2023 |
Event | 11th International Conference on Learning Representations, ICLR 2023 - Kigali, Rwanda Duration: May 1 2023 → May 5 2023 |
Conference
Conference | 11th International Conference on Learning Representations, ICLR 2023 |
---|---|
Country/Territory | Rwanda |
City | Kigali |
Period | 5/1/23 → 5/5/23 |
All Science Journal Classification (ASJC) codes
- Language and Linguistics
- Computer Science Applications
- Education
- Linguistics and Language