Truncating Temporal Differences: On the Efficient Implementation of TD(lambda) for Reinforcement Learning

Open in new window