Learning from Human-Generated Reward

September 2012

The Research Question: How and to what extent can agents harness the information contained in human-generated signals of reward to learn sequential decision-making tasks?