On the Opportunities and Challenges of Offline Reinforcement Learning for Recommender Systems.
Xiaocong ChenSiyu WangJulian J. McAuleyDietmar JannachLina YaoPublished in: CoRR (2023)
Keyphrases
- recommender systems
- reinforcement learning
- collaborative filtering
- lessons learned
- real time
- user preferences
- real world
- learning process
- markov decision processes
- matrix factorization
- key issues
- function approximation
- emerging technologies
- information technology
- supervised learning
- learning algorithm
- user profiling
- reinforcement learning algorithms
- trust aware
- robotic control
- technical challenges
- temporal difference
- model free
- user modeling
- information overload
- optimal control
- machine learning
- optimal policy
- multi agent