Title | ||
---|---|---|
Ramp-loss nonparallel support vector regression: Robust, sparse and scalable approximation. |
Abstract | ||
---|---|---|
Although the twin support vector regression (TSVR) has been extensively studied and diverse variants are successfully developed, when it comes to outlier-involved training set, the regression model can be wrongly driven towards the outlier points, yielding extremely poor generalization performance. To overcome such shortcoming, a Ramp-loss nonparallel support vector regression (RL-NPSVR) is proposed in this work. By adopting Ramp ε-insensitive loss function and another Ramp-type linear loss function, RL-NPSVR can not only explicitly filter noise and outlier suppression but also have an excellent sparseness. The non- convexity of RL-NPSVR is solved by concave–convex programming (CCCP). Because a regularized term is added into each primal problem by rigidly following the structural risk minimization (SRM) principle, CCCP actually solves a series of reconstructed convex optimizations which have the same formulation of dual problem as the standard SVR, so that computing inverse matrix is avoided and SMO-type fast algorithm can be used to accelerate the training process. Numerical experiments on various datasets have verified the effectiveness of our proposed RL-NPSVR in terms of outlier sensitivity, generalization ability, sparseness and scalability. |
Year | DOI | Venue |
---|---|---|
2018 | 10.1016/j.knosys.2018.02.016 | Knowledge-Based Systems |
Keywords | Field | DocType |
Support vector regression,Twin support vector regression,Ramp loss,CCCP,Sparseness | Data mining,Convexity,Regression analysis,Computer science,Matrix (mathematics),Support vector machine,Outlier,Algorithm,Duality (optimization),Structural risk minimization,Scalability | Journal |
Volume | ISSN | Citations |
147 | 0950-7051 | 2 |
PageRank | References | Authors |
0.36 | 34 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Long Tang | 1 | 7 | 1.44 |
Ying-Jie Tian | 2 | 18 | 6.34 |
Chunyan Yang | 3 | 4 | 1.74 |
Panos M. Pardalos | 4 | 141 | 19.60 |