Wavelet Kernel Matching Pursuit Machine

Abstract

Kernel Matching Pursuit Machine is a relatively new learning algorithm utilizing Mercer kernels to produce non-linear version of conventional supervised and unsupervised learning algorithm. But the commonly used Mercer kernels can’t expand a set of complete bases in the feature space (subspace of the square and integrable space). Hence the decision-function found by the machine can’t approximate arbitrary objective function in feature space as precise as possible. Wavelet technique shows promise for both nonstationary signal approximation and classification, so we combine KMPM with wavelet technique to improve the performance of the machine, and put forward a wavelet translation invariant kernel, which is a Mercer admissive kernel by theoretical analysis. The wavelet kernel matching pursuit machine is constructed in this paper by a translation-invariant wavelet kernel. It is shown that WKMPM is much more effective in the problems of regression and pattern recognition by the number of comparable experiments.