RL4RS: A Real-World Benchmark for Reinforcement Learning based Recommender System.
Kai WangZhene ZouQilin DengYue ShangMinghao ZhaoRunze WuXudong ShenTangjie LyuChangjie FanPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- real world
- recommender systems
- function approximation
- reinforcement learning algorithms
- control problems
- state space
- markov decision processes
- wide range
- learning algorithm
- optimal policy
- multi agent
- collaborative filtering
- synthetic data
- optimal control
- temporal difference learning
- machine learning
- policy search
- user preferences
- rl algorithms
- temporal difference
- supervised learning
- model free
- real robot
- continuous state
- continuous state and action spaces
- direct policy search
- user profiles
- partially observable
- learning capabilities
- action space
- control policy
- action selection
- information filtering
- transfer learning
- policy gradient
- exploration exploitation
- partially observable domains
- learning process
- data sets