Trust Region

Overall Progress 0%

TRPO constrained optimization and natural gradient; KL constraint.

Go to Chapter 42: Trust Region Policy Optimization (TRPO) →