Fast Kernel Learning using Sequential Minimal Optimization

While classical kernel-based classifiers are based on a single kernel, in practice it is often desirable to base classifiers on combinations of multiple kernels. Lanckriet et al. (2004) considered conic combinations of kernel matrices for the support vector machine (SVM), and showed that the optimization of the coefficients of such a combination reduces to a convex optimization problem known as a quadratically-constrained quadratic program (QCQP). Unfortunately, current convex optimization toolboxes can solve this problem only for a small number of kernels and a small number of data points; moreover, the sequential minimal optimization (SMO) techniques that are essential in large-scale implementations of the SVM cannot be applied because the cost function is non-differentiable. We propose a novel dual formulation of the QCQP as a second-order cone programming problem, and show how to exploit the technique of Moreau-Yosida regularization to yield a formulation to which SMO techniques can be applied. We present experimental results that show that our SMO-based algorithm is significantly more efficient than the general-purpose interior point methods available in current optimization toolboxes.

[1]  D. Anderson,et al.  Algorithms for minimization without derivatives , 1974 .

[2]  J. Hiriart-Urruty,et al.  Convex analysis and minimization algorithms , 1993 .

[3]  Dimitri P. Bertsekas,et al.  Nonlinear Programming , 1995 .

[4]  Claude Lemaréchal,et al.  Practical Aspects of the Moreau-Yosida Regularization: Theoretical Preliminaries , 1997, SIAM J. Optim..

[5]  Paul S. Bradley,et al.  Feature Selection via Concave Minimization and Support Vector Machines , 1998, ICML.

[6]  Thorsten Joachims,et al.  Making large-scale support vector machine learning practical , 1999 .

[7]  Stephen P. Boyd,et al.  Applications of second-order cone programming , 1998 .

[8]  John C. Platt,et al.  Fast training of support vector machines using sequential minimal optimization, advances in kernel methods , 1999 .

[9]  B. Schölkopf,et al.  Advances in kernel methods: support vector learning , 1999 .

[10]  Knud D. Andersen,et al.  The Mosek Interior Point Optimizer for Linear Programming: An Implementation of the Homogeneous Algorithm , 2000 .

[11]  S. Sathiya Keerthi,et al.  Improvements to Platt's SMO Algorithm for SVM Classifier Design , 2001, Neural Computation.

[12]  Yves Grandvalet,et al.  Adaptive Scaling for Feature Selection in SVMs , 2002, NIPS.

[13]  Alexander J. Smola,et al.  Hyperkernels , 2002, NIPS.

[14]  Nello Cristianini,et al.  Learning the Kernel Matrix with Semidefinite Programming , 2002, J. Mach. Learn. Res..

[15]  Sayan Mukherjee,et al.  Choosing Multiple Parameters for Support Vector Machines , 2002, Machine Learning.