replay

Deep Q-learning(DQN) – Experience Replay

On: July 19, 2019

In: AI, Deep Machine Learning, Reinforcement Learning

To perform experience replay we store the agent’s experiences et=(st,at,rt,st+1) Then we use a random sample of these prior actions instead of the most recent action to proceed. * This removes correlations in the observation sequence and smooths changes in the data distribution. * Iterative update adjusts Q towards targetContinue Reading