Nonlinear classifiers for ranking problems based on kernelized SVM
This work addresses ranking and top-accuracy problems in machine learning, but it is incremental as it builds on a previously established linear framework.
The paper tackles the problem of extending a linear classification framework for ranking and top-accuracy tasks to nonlinear classifiers, achieving this by dualizing the problems, incorporating kernels, and proposing a componentwise dual ascent method.
Many classification problems focus on maximizing the performance only on the samples with the highest relevance instead of all samples. As an example, we can mention ranking problems, accuracy at the top or search engines where only the top few queries matter. In our previous work, we derived a general framework including several classes of these linear classification problems. In this paper, we extend the framework to nonlinear classifiers. Utilizing a similarity to SVM, we dualize the problems, add kernels and propose a componentwise dual ascent method.