03ReinforcementLearning2.6, n-step TD methods
From Wulfram Gerstner
views
comments
From Wulfram Gerstner
Normal TD methods compare states that are nearest neighbors whereas n-step methods compare states that are n steps apart.