Asynchronous Parallel Empirical Variance Guided Algorithms for the Thresholding Bandit Problem.
Jie ZhongYijun HuangJi LiuPublished in: CoRR (2017)
Keyphrases
- theoretical analysis
- computational cost
- learning algorithm
- computationally efficient
- computational complexity
- data structure
- data streams
- active learning
- significant improvement
- bayesian networks
- computer vision
- machine learning
- gray level
- benchmark datasets
- orders of magnitude
- times faster
- parallel processors
- variance reduction
- parallel version