RL Unplugged: Benchmarks for Offline Reinforcement Learning.
Çaglar GülçehreZiyu WangAlexander NovikovTom Le PaineSergio Gómez ColmenarejoKonrad ZolnaRishabh AgarwalJosh MerelDaniel J. MankowitzCosmin PaduraruGabriel Dulac-ArnoldJerry LiMohammad NorouziMatt HoffmanOfir NachumGeorge TuckerNicolas HeessNando de FreitasPublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- function approximation
- state space
- rl algorithms
- model free
- reinforcement learning algorithms
- temporal difference learning
- optimal policy
- real time
- transfer learning
- reinforcement learning agents
- autonomous learning
- multi agent
- supervised learning
- markov decision processes
- learning problems
- continuous state
- control problems
- action space
- learning algorithm
- reinforcement learning methods
- policy search
- benchmark suite
- machine learning
- temporal difference
- control strategies
- learning process
- actor critic
- optimal control
- dynamic programming
- exploration exploitation
- direct policy search