A Best-of-Both-Worlds Algorithm for Bandits with Delayed Feedback.
Saeed MasoudianJulian ZimmertYevgeny SeldinPublished in: NeurIPS (2022)
Keyphrases
- dynamic programming
- optimization algorithm
- learning algorithm
- similarity measure
- single pass
- clustering method
- particle swarm optimization
- high accuracy
- probabilistic model
- experimental evaluation
- computational complexity
- np hard
- computationally efficient
- genetic algorithm
- times faster
- matching algorithm
- experimental study
- objective function
- expectation maximization
- optimal solution
- cost function
- online learning
- linear programming
- simulated annealing
- detection algorithm
- feature selection
- k means
- classification algorithm
- computational cost
- significant improvement
- convergence rate
- recognition algorithm
- improved algorithm