Optimizing Long-term Value for Auction-Based Recommender Systems via On-Policy Reinforcement Learning.
Ruiyang XuJalaj BhandariDmytro KorenkevychFan LiuYuchen HeAlex NikulkovZheqing ZhuPublished in: CoRR (2023)
Keyphrases
- recommender systems
- long term
- reinforcement learning
- optimal policy
- policy search
- short term
- exploration exploitation tradeoff
- markov decision process
- collaborative filtering
- action selection
- partially observable environments
- markov decision processes
- reinforcement learning problems
- policy gradient
- policy iteration
- function approximation
- state space
- reinforcement learning algorithms
- actor critic
- partially observable
- action space
- control policy
- function approximators
- reward function
- information overload
- markov decision problems
- model free
- user preferences
- partially observable markov decision processes
- user modeling
- policy gradient methods
- online auctions
- user profiling
- state action
- temporal difference
- state and action spaces
- policy evaluation
- user profiles
- machine learning
- continuous state spaces
- learning algorithm
- approximate dynamic programming
- continuous state
- control policies
- rl algorithms
- decision problems
- bidding strategies
- personalized recommendation
- infinite horizon
- partially observable domains
- combinatorial auctions
- reinforcement learning methods
- cold start problem
- mechanism design
- long run
- recommendation systems
- agent receives