Time-Scale Separation in Q-Learning: Extending TD($\triangle$) for Action-Value Function Decomposition

Open in new window