Best-of-three-worlds Analysis for Linear Bandits with Follow-the-regularized-leader Algorithm.
Fang KongCanzhe ZhaoShuai LiPublished in: COLT (2023)
Keyphrases
- learning algorithm
- detection algorithm
- regularized least squares
- computational complexity
- optimization algorithm
- worst case
- np hard
- dynamic programming
- high accuracy
- improved algorithm
- input data
- experimental evaluation
- k means
- probabilistic model
- cost function
- linear complexity
- linear systems
- theoretical analysis
- path planning
- data analysis
- solution path
- objective function
- linear model
- closed form
- preprocessing
- expectation maximization
- particle swarm optimization
- significant improvement
- neural network