Cold-Start Reinforcement Learning with Softmax Policy Gradient.
Nan DingRadu SoricutPublished in: NIPS (2017)
Keyphrases
- policy gradient
- cold start
- reinforcement learning
- recommender systems
- actor critic
- function approximation
- temporal difference learning
- reinforcement learning algorithms
- collaborative filtering
- data sparsity
- policy search
- policy gradient methods
- optimal control
- tag recommendation
- user preferences
- gradient method
- approximation methods
- reinforcement learning methods
- control problems
- function approximators
- state space
- model free
- matrix factorization
- state action
- variance reduction
- implicit feedback
- user ratings
- personalized recommendation
- transfer learning
- temporal difference
- user profiles
- multi agent
- learning problems
- prediction accuracy
- text categorization
- dynamic programming