Mixing Human Demonstrations with Self-Exploration in Experience Replay for Deep Reinforcement Learning.

Dylan Klein Akansel Cosgun

Published in: CoRR (2021)

Keyphrases

reinforcement learning
guided exploration
action selection
machine learning
state space
optimal policy
function approximation
human interaction
deep learning
active exploration
human subjects
autonomous learning
exploration exploitation
model based reinforcement learning
data sets
dynamic programming
multi agent