Worst-Case Absolute Loss Bounds for Linear Learning Algorithms.
Tom BylanderPublished in: AAAI/IAAI (1997)
Keyphrases
- loss bounds
- worst case
- square loss
- learning algorithm
- expert advice
- target concept
- upper bound
- linear regression
- temporal difference learning
- lower bound
- bregman divergences
- machine learning algorithms
- np hard
- learning tasks
- machine learning
- aggregating algorithm
- closed form
- learning problems
- sample size
- training examples
- training data
- multi class
- special case
- uniform distribution
- maximum likelihood
- computational complexity