DeepAveragers: Offline Reinforcement Learning by Solving Derived Non-Parametric MDPs.
Aayam ShresthaStefan LeePrasad TadepalliAlan FernPublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- markov decision processes
- markov decision problems
- sequential decision making problems
- optimal policy
- semi markov decision processes
- state space
- machine learning
- function approximation
- policy iteration
- partially observable
- policy search
- average reward
- state and action spaces
- continuous state
- reward function
- reinforcement learning agents
- continuous state and action spaces
- dynamic programming
- action space
- learning algorithm
- approximate dynamic programming
- markov decision process
- action selection
- model free
- model based reinforcement learning
- transition model
- stochastic shortest path
- continuous state spaces
- rl algorithms
- control problems
- reinforcement learning algorithms
- temporal difference
- solving problems
- finite state
- probability density function
- utility function
- search algorithm