Generalized Policy Iteration
Generalized Policy Iteration (GPI) is a classical framework in reinforcement learning that characterizes policy learning as an alternating cycle of policy evaluation and policy improvement, where both the improvement mechanism and the evaluation criterion are externally given and fixed.
Timeline (last 90 days)
Generalized Policy Iteration (GPI) characterizes policy learning as a cycle of alternating policy evaluation and policy improvement.
The implicit premise of GPI is that the update principle and the evaluation benchmark are both external to the agent, externally given and fixed.
All Facts (2)
Generalized Policy Iteration (GPI) characterizes policy learning as a cycle of alternating policy evaluation and policy improvement.
50%UnverifiedThe implicit premise of GPI is that the update principle and the evaluation benchmark are both external to the agent, externally given and fixed.
50%