Multi-User Reinforcement Learning with Low Rank Rewards.
Dheeraj Mysore NagarajSuhas S. KowshikNaman AgarwalPraneeth NetrapalliPrateek JainPublished in: ICML (2023)
Keyphrases
- multi user
- low rank
- reinforcement learning
- linear combination
- convex optimization
- matrix factorization
- missing data
- virtual environment
- low rank matrix
- high order
- markov decision processes
- matrix completion
- semi supervised
- state space
- user interface
- virtual world
- singular value decomposition
- function approximation
- augmented reality
- reinforcement learning algorithms
- rank minimization
- matrix decomposition
- multi granularity
- kernel matrix
- single user
- multiple users
- high dimensional data
- optimal policy
- singular values
- non rigid structure from motion
- robust principal component analysis
- reward function
- low rank matrices
- learning algorithm
- model free
- learning problems
- temporal difference
- learning process
- transfer learning
- minimization problems
- expert systems
- trace norm
- machine learning
- pairwise
- regularized regression
- pattern recognition
- high dimensional
- dynamic programming
- collaborative filtering
- higher order
- natural images