In this lecture, we will take our first step outside binary classification towards more complex outputs. We will consider the case of multiclass classification – where the classifier has to choose from among a finite (and possibly small) number of inference outcomes.
Lecture and readings
- Lecture slides
- Thomas G. Dietterich and Ghulum Bakiri, Solving Multiclass Learning Problems via Error-Correcting Output Codes, JAIR, 1995.
- Erin L. Allwein, Robert E. Schapire, Yoram Singer, Reducing Multiclass to Binary: A Unifying Approach for Margin Classifiers, ICML 2000. (Slides from ICML 2010)
- Koby Crammer and Yoram Singer, On the Algorithmic Implementation of Multiclass Kernel-based Vector Machines, JMLR 2001.
- Weston, Jason, and Chris Watkins. Multi-class support vector machines. Technical Report CSD-TR-98-04, Department of Computer Science, Royal Holloway, University of London, May, 1998.
- Sariel Har-Peled, Dan Roth and Dav Zimak, Constraint Classification for Multiclass Classification and Ranking, NIPS 2002.