Optimal Selection of Time-Frequency Representations for Signal Classification: a Kernel-Target Alignment Approach

In this paper, we propose a method for selecting time-frequency distributions appropriate for given learning tasks. It is based on a criterion that has recently emerged from the machine learning literature: the kernel-target alignment. This criterion makes possible to find the optimal representation for a given classification problem without designing the classifier itself. Some possible applications of our framework are discussed. The first one provides a computationally attractive way of adjusting the free parameters of a distribution to improve classification performance. The second one is related to the selection, from a set of candidates, of the distribution that best facilitates a classification task. The last one addresses the problem of optimally combining several distributions

[1]  N. Cristianini,et al.  On Kernel-Target Alignment , 2001, NIPS.

[2]  Paulo Gonçalves,et al.  Adaptive diffusion as a versatile tool for time-frequency and time-scale representations processing: a review , 2005, IEEE Transactions on Signal Processing.

[3]  C. Richard,et al.  Kernel machines : une nouvelle méthode pour l'optimisation de l'alignement des noyaux et l'amélioration des performances , 2005 .

[4]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[5]  Jasha Droppo,et al.  Optimizing time-frequency distributions for automatic classification , 1997, Optics & Photonics.

[6]  Bernhard Schölkopf,et al.  A Generalized Representer Theorem , 2001, COLT/EuroCOLT.

[7]  Edward Y. Chang,et al.  Formulating distance functions via the kernel trick , 2005, KDD '05.

[8]  Patrick Flandrin,et al.  Improving the readability of time-frequency and time-scale representations by the reassignment method , 1995, IEEE Trans. Signal Process..

[9]  Nello Cristianini,et al.  On the Extensions of Kernel Alignment , 2002 .

[10]  N. Aronszajn.  Theory of Reproducing Kernels. , 1950 .

[11]  Gunnar Rätsch,et al.  An introduction to kernel-based learning algorithms , 2001, IEEE Trans. Neural Networks.

[12]  C. Richard,et al.  Reconnaissance des formes par méthodes à noyau dans le domaine temps-fréquence , 2005 .

[13]  Douglas L. Jones,et al.  An adaptive optimal-kernel time-frequency representation , 1993, 1993 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[14]  P.J.W. Rayner,et al.  Optimized support vector machines for nonstationary signal classification , 2002, IEEE Signal Processing Letters.

[15]  N. Cristianini,et al.  Optimizing Kernel Alignment over Combinations of Kernel , 2002 .

[16]  C. Doncarli,et al.  Improved optimization of time-frequency-based signal classifiers , 2001, IEEE Signal Processing Letters.