Intrinsic Rewards for Reinforcement Learning Within Complex 2D Environments.
Nathaniel GrabaskasZhizhen WangPublished in: IntelliSys (1) (2021)
Keyphrases
- reinforcement learning
- markov decision processes
- real world
- temporal difference
- function approximation
- highly dynamic
- state space
- complex environments
- temporal abstractions
- reward function
- complex systems
- dynamic environments
- learning algorithm
- machine learning
- decision problems
- optimal policy
- multi agent
- function approximators
- data sets