pyrobolearn/pyrobolearn/returns

Returns / Estimators

This folder provides the various returns / estimators used in reinforcement learning (in the evaluation step of RL algorithms), such as:

• Total reward estimator
• Action reward estimator
• Baseline reward estimator
• State value estimator: V(s)
• State-action value estimator: Q(s,a)
• Advantage estimator: A(s,a) = Q(s,a) - V(s)
• TD residual estimator
• Generalized advantage estimator (GAE)

what to look/check next?

Have a look at algos/rl_algos, storages and values folders.