On the Dual Formulation of Boosting Algorithms.
Chunhua ShenHanxi LiPublished in: IEEE Trans. Pattern Anal. Mach. Intell. (2010)
Keyphrases
- dual formulation
- boosting algorithms
- multi class
- loss function
- cost sensitive
- total variation
- convex optimization
- weak learners
- primal dual
- binary classification
- weak classifiers
- iterative algorithms
- generalization error
- learning algorithm
- ensemble methods
- base classifiers
- image restoration
- pairwise
- benchmark datasets
- special case
- machine learning
- learning tasks
- sample size
- active learning
- misclassification costs
- training data
- higher order
- object detection