A Simple Multi-Class Boosting Framework with Theoretical Guarantees and Empirical Proficiency.
Ron AppelPietro PeronaPublished in: ICML (2017)
Keyphrases
- multi class
- theoretical guarantees
- boosting framework
- boosting algorithms
- multi class classification
- binary classification
- pairwise
- support vector machine
- feature selection
- cost sensitive
- machine learning
- object detection
- loss function
- binary classifiers
- worst case
- nearest neighbor
- optimal solution
- markov random field
- semi supervised learning
- base classifiers
- active learning
- training set
- training data