An Improved Parametrization and Analysis of the EXP3++ Algorithm for Stochastic and Adversarial Bandits.
Yevgeny SeldinGábor LugosiPublished in: COLT (2017)
Keyphrases
- detection algorithm
- experimental evaluation
- convergence rate
- times faster
- artificial neural networks
- dynamic programming
- expectation maximization
- learning algorithm
- decision trees
- objective function
- monte carlo
- cost function
- neural network
- improved algorithm
- computationally efficient
- np hard
- significant improvement
- k means
- preprocessing
- data sets
- probabilistic model
- worst case
- online learning
- computational cost
- theoretical analysis
- segmentation algorithm
- optimization algorithm
- clustering method
- matching algorithm
- data analysis