Bandit-Based Algorithms for Budgeted Learning.
Kun DengChris BourkeStephen ScottJulie SundermanYaling ZhengPublished in: ICDM (2007)
Keyphrases
- learning algorithm
- learning systems
- learning tasks
- learning process
- benchmark datasets
- reinforcement learning
- theoretical analysis
- automatically learned
- noise tolerant
- data sets
- worst case
- supervised learning
- multi class
- pairwise
- learning styles
- computational cost
- prior knowledge
- background knowledge
- learning problems
- average case
- online algorithms
- computational complexity
- data structure
- bandit problems