Abstract
Boosting combines weak classifiers to form highly accurate predictors. Although the case of binary classification is well understood, in the multiclass setting, the "correct" requirements on the weak classifier, or the notion of the most efficient boosting algorithms are missing. In this paper, we create a broad and general framework, within which we make precise and identify the optimal requirements on the weak-classifier, as well as design the most effective, in a certain sense, boosting algorithms that assume such requirements.
Original language | English (US) |
---|---|
Pages (from-to) | 437-497 |
Number of pages | 61 |
Journal | Journal of Machine Learning Research |
Volume | 14 |
Issue number | 1 |
State | Published - Feb 1 2013 |
All Science Journal Classification (ASJC) codes
- Software
- Control and Systems Engineering
- Statistics and Probability
- Artificial Intelligence
Keywords
- Boosting
- Drifting games
- Multiclass
- Weak learning condition