Improved Algorithms for Stochastic Linear Bandits Using Tail Bounds for Martingale Mixtures.
Hamish FlynnDavid ReebMelih KandemirJan R. PetersPublished in: NeurIPS (2023)
Keyphrases
- regret bounds
- learning algorithm
- theoretical analysis
- worst case
- change detection
- generalization bounds
- linear space
- error bounds
- upper and lower bounds
- generalization error bounds
- multi armed bandit
- mistake bound
- theoretical guarantees
- average case
- linear models
- model selection
- optimization problems
- significant improvement
- lower bound
- computational complexity
- image segmentation