Are Deep Policy Gradient Algorithms Truly Policy Gradient Algorithms?

11/06/2018
by   Andrew Ilyas, et al.
4

We study how the behavior of deep policy gradient algorithms reflects the conceptual framework motivating their development. We propose a fine-grained analysis of state-of-the-art methods based on key aspects of this framework: gradient estimation, value prediction, optimization landscapes, and trust region enforcement. We find that from this perspective, the behavior of deep policy gradient algorithms often deviates from what their motivating framework would predict. Our analysis suggests first steps towards solidifying the foundations of these algorithms, and in particular indicates that we may need to move beyond the current benchmark-centric evaluation methodology.

READ FULL TEXT

Please sign up or login with your details

Forgot password? Click here to reset