Contextual Bandits with Packing and Covering Constraints: A Modular Lagrangian Approach via Regression.
Aleksandrs SlivkinsKarthik Abinav SankararamanDylan J. FosterPublished in: COLT (2023)
Keyphrases
- constraint satisfaction
- regression model
- constrained optimization
- context sensitive
- contextual information
- derivation rules
- support vector regression
- dynamic programming
- support vector
- linear regression
- integer programming
- optimal solution
- geometric constraints
- reinforcement learning
- linear constraints
- lagrange multipliers
- regression function
- stochastic systems
- neural network