Non-exponential Reward Discounting in Reinforcement Learning.
Raja Farrukh AliPublished in: AAAI (2023)
Keyphrases
- reinforcement learning
- function approximation
- state space
- eligibility traces
- learning algorithm
- reinforcement learning algorithms
- reward function
- model free
- markov decision processes
- dynamic programming
- optimal policy
- learning agent
- temporal difference
- supervised learning
- multi agent
- learning problems
- total reward
- average reward
- policy evaluation
- partially observable environments
- robotic control
- data sets
- expected reward
- multi agent reinforcement learning
- temporal difference learning
- linear complexity
- action space
- robot control
- learning capabilities
- action selection