Abstract | ||
---|---|---|
This paper presents a decomposition method for efficiently constructing ℓ1-norm Support Vector Machines (SVMs). The decomposition algorithm introduced in this paper possesses many desirable properties. For example, it is provably convergent, scales well to large datasets, is easy to implement, and can be extended to handle support vector regression and other SVM variants. We demonstrate the efficiency of our algorithm by training on (dense) synthetic datasets of sizes up to 20 million points (in ℝ32). The results show our algorithm to be several orders of magnitude faster than a previously published method for the same task. We also present experimental results on real data sets—our method is seen to be not only very fast, but also highly competitive against the leading SVM implementations. |
Year | DOI | Venue |
---|---|---|
2006 | 10.1007/11871842_78 | ECML |
Keywords | Field | DocType |
efficient large scale,leading svm implementation,decomposition algorithm,decomposition method,large datasets,desirable property,million point,vector machines,1-norm support,synthetic datasets,linear programming support vector,svm variant,support vector,linear program,support vector machine | Data mining,Orders of magnitude (numbers),Regression analysis,Support vector machine,Linear machine,Algorithm,Decomposition method (constraint satisfaction),Implementation,Linear programming,Mathematics,Statistical analysis | Conference |
Volume | ISSN | ISBN |
4212 | 0302-9743 | 3-540-45375-X |
Citations | PageRank | References |
7 | 0.61 | 7 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
suvrit sra | 1 | 14 | 1.20 |
furnkranz | 2 | 11 | 1.14 |
Tobias Scheffer | 3 | 1862 | 139.64 |
Myra Spiliopoulou | 4 | 2297 | 232.72 |