Tsallis-INF: An Optimal Algorithm for Stochastic and Adversarial Bandits.
Julian ZimmertYevgeny SeldinPublished in: J. Mach. Learn. Res. (2021)
Keyphrases
- dynamic programming
- optimal solution
- optimization algorithm
- high accuracy
- preprocessing
- experimental evaluation
- detection algorithm
- times faster
- globally optimal
- expectation maximization
- worst case
- computational cost
- np hard
- search space
- locally optimal
- weighting coefficients
- information theory
- matching algorithm
- monte carlo
- decision trees
- cost function
- k means
- learning algorithm
- online learning
- tree structure
- input data
- linear programming
- improved algorithm
- computational complexity
- optimal strategy
- multi agent
- operating point