Login / Signup
Action Guidance: Getting the Best of Sparse Rewards and Shaped Rewards for Real-time Strategy Games.
Shengyi Huang
Santiago Ontañón
Published in:
CoRR (2020)
Keyphrases
</>
real time strategy games
reinforcement learning
multiarmed bandit
markov decision processes
bandit problems
reward shaping
expected reward
discounted reward
domain knowledge
multi agent
special case
optimal policy
case based planning