Contexts can be Cheap: Solving Stochastic Contextual Bandits with Linear Bandit Algorithms.
Osama A. HannaLin F. YangChristina FragouliPublished in: CoRR (2022)
Keyphrases
- multi armed bandit
- regret bounds
- computational cost
- learning algorithm
- combinatorial optimization
- orders of magnitude
- significant improvement
- low cost
- stochastic search
- genetic algorithm
- worst case
- computationally efficient
- theoretical analysis
- active learning
- data structure
- reinforcement learning
- backtracking algorithm