Abstract | ||
---|---|---|
The success of kernel methods is very much dependent on the choice of kernels. Multiple kernel learning (MKL) aims at learning a combination of different kernels in order to better match the underlying problem instead of using a single fixed kernel. In this paper, we propose a simple but effective multiclass MKL method by a two-stage strategy, in which the first stage finds the kernel weights to combine the kernels, and the second stage trains a standard multiclass support vector machine (SVM). Specifically, we first present an evaluation criterion named multiclass kernel polarization (MKP) to assess the quality of a kernel in the multiclass classification scenario, and then develop a heuristic rule to directly assign a weight to each kernel based on the quality of the individual kernel. MKP is a multiclass extension of the kernel polarization, which is a universal kernel evaluation criterion for kernel design and learning. Comprehensive experiments are conducted on several UCI benchmark examples and the results well demonstrate the effectiveness and efficiency of our approach. |
Year | DOI | Venue |
---|---|---|
2013 | 10.1016/j.knosys.2013.04.006 | Knowledge-Based Systems |
Keywords | Field | DocType |
Multiple kernel learning (MKL),Multiclass kernel polarization,Support vector machine (SVM),Multiclass classification,Model selection | Radial basis function kernel,Pattern recognition,Kernel embedding of distributions,Computer science,Multiple kernel learning,Tree kernel,Polynomial kernel,Artificial intelligence,Kernel method,Variable kernel density estimation,Machine learning,Multiclass classification | Journal |
Volume | Issue | ISSN |
48 | null | 0950-7051 |
Citations | PageRank | References |
14 | 0.50 | 26 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Tinghua Wang | 1 | 44 | 2.42 |
Dongyan Zhao | 2 | 998 | 96.35 |
Yansong Feng | 3 | 735 | 64.17 |