- Title
- Boosting support vector machines successfully
- Creator
- Ting, Kaiming; Zhu, Lian
- Date
- 2009
- Type
- Text; Conference paper
- Identifier
- http://researchonline.federation.edu.au/vital/access/HandleResolver/1959.17/75244
- Identifier
- vital:7320
- Identifier
-
https://doi.org/10.1007/978-3-642-02326-2_51
- Identifier
- ISBN:0302-9743
- Abstract
- Boosting has been shown to improve the predictive performance of unstable learners such as decision trees, but not of stable learners like support vector machines (SVM). In addition to the model stability problem, the high computational cost of SVM prohibits it from generating multiple models to form an ensemble for large data sets. This paper introduces a method that not only enables boosting to improve the predictive performance of SVM, but also reduces the computational cost to make ensembles of SVM feasible for large data sets. The method proposes to build local models, instead of global models; and it is the first method, to the best of our knowledge, to solve the two problems in boosting SVM at the same time. The proposed method to boost SVM also performs better than boosting decision trees in term of predictive accuracy in our experiments
- Publisher
- Springer-verlag
- Relation
- 8th International Workshop on Multiple Classifier Systems (MCS 2009) p. 509-518
- Rights
- This metadata is freely available under a CCO license
- Subject
- Boosting; Support vector machines; Local model; 0801 Artificial Intelligence and Image Processing
- Reviewed
- Hits: 471
- Visitors: 473
- Downloads: 0
| Thumbnail | File | Description | Size | Format |
|---|