Approximate Difference Rewards for Scalable Multigent Reinforcement Learning.
Arambam James SinghAkshat KumarHoong Chuin LauPublished in: AAMAS (2021)
Keyphrases
- reinforcement learning
- function approximation
- policy evaluation
- markov decision processes
- state space
- reinforcement learning algorithms
- learning algorithm
- reward shaping
- temporal difference
- model free
- optimal policy
- reward function
- continuous state
- machine learning
- highly scalable
- reinforcement learning methods
- learning process
- web scale
- multiarmed bandit
- exact solution
- transfer learning
- lower bound
- convex functions
- complex domains
- markov decision process
- policy iteration
- rl algorithms
- dynamic programming
- neural network
- data sets