Elaine Schaertl Short

Policy Shaping with Supervisory Attention Driven Exploration

2018

Robots deployed for long periods of time need to be able to explore and learn from their environment. One approach to this problem has been reinforcement learning (RL), in which robots receive rewards from the …