Support vector machines (SVMs), though accurate, are not preferred in applications requiring great classification speed, due to the number of support vectors being large. To overcome this problem we devise a primal method with the following properties: (1) it decouples the idea of basis functions from the concept of support vectors; (2) it greedily finds a set of kernel basis functions of a specified maximum size (dmax) to approximate the SVM primal cost function well; (3) it is efficient and roughly scales as O(nd 2 max) where n is the number of training examples; and, (4) the number of basis functions it requires to achieve an accuracy close to the SVM accuracy is usually far less than the number of SVM support vectors.
|
697
|
Making large-Scale SVM Learning Practical
– Joachims
- 1999
|
|
401
|
Matching pursuits with timefrequency dictionaries
– Mallat, Zhang
- 1993
|
|
204
|
Sparse Bayesian learning and the relevance vector machine
– Tipping
|
|
196
|
der Vorst. An iterative solution method for linear systems of which the coefficient matrix is a M-matrix
– Meijerink, Van
- 1977
|
|
176
|
Greedy function approximation: a gradient boosting machine
– Friedman
- 2001
|
|
129
|
Sequenital minimal optimization: A fast algorithm for training support vector machines
– Platt
- 1998
|
|
94
|
Sparse greedy matrix approximation for machine learning
– Smola, Schökopf
- 2000
|
|
91
|
Improving the accuracy and speed of support vector machines
– Burges, Scholkopf
- 1997
|
|
91
|
Input space vs. feature space in kernel-based methods
– Scholkopf, Mika, et al.
- 1999
|
|
73
|
Training invariant support vector machines
– DeCoste, Schölkopf
- 2001
|
|
69
|
RSVM: Reduced support vector machines
– Lee, Mangasarian
|
|
54
|
Sparse greedy gaussian process regression
– Smola, Bartlett
- 2001
|
|
50
|
Fast sparse gaussian process methods: The informative vector machine
– Lawrence, Seeger, et al.
- 2002
|
|
35
|
Reducing the run-time complexity of support vector machines
– Osuna, Girosi
- 1998
|
|
26
|
Exact simplification of support vector solutions
– Downs, Gates, et al.
|
|
26
|
Fast forward selection to speed up sparse gaussian process regression
– Seeger, Williams, et al.
- 2003
|
|
23
|
Kernel matching pursuit
– Vincent, Bengio
- 2002
|
|
20
|
A modified finite newton method for fast solution of large scale linear svms
– Keerthi, DeCoste
- 2005
|
|
18
|
A study on reduced support vector machines
– Lin, Lin
- 2003
|
|
18
|
Robust Boosting via Convex Optimization
– Rätsch
- 2001
|
|
16
|
Mark: a boosting algorithm for heterogeneous kernel models
– Bennett, Momma, et al.
- 2002
|
|
12
|
Comparison of basis selection methods
– Adler, Rao, et al.
- 1996
|
|
12
|
Training a support vector machine in the primal
– Chapelle
- 2006
|
|
8
|
Predictive low-rank decomposition for kernel methods
– Bach, Jordan
- 2005
|
|
8
|
Column-generation boosting methods for mixture of kernels
– Bi, Zhang, et al.
- 2004
|
|
8
|
A finite newton method for classification
– Mangasarian
- 2002
|
|
6
|
A matching pursuit approach to sparse Gaussian process regression
– Keerthi, Chu
- 2006
|
|
4
|
Sparseness of support vector machines – some asymptotically sharp bounds
– Steinwart
- 2004
|
|
3
|
Growing support vector classifiers with controlled complexity
– Parrado-Hernandez, Mora-Jimenez, et al.
- 2003
|
|
2
|
Low rank updates for the Cholesky decomposition
– Seeger
- 2004
|
|
1
|
Optimal reduced-set vectors for support vector machines with a quadratic kernel
– Thies, Weber
|
|
1
|
Benchmark repository
– Rätsch
|
|
1
|
space vs. feature space in kernel-based methods
– Input
- 1999
|
