A Kernel Approach for Vector Quantization with Guaranteed Distortion Bounds

Michael E. Tipping, Bernhard Schölkopf
Proceedings of the Eighth International Workshop on Artificial Intelligence and Statistics, PMLR R3:298-303, 2001.

Abstract

We propose a kernel method for vector quantization and clustering. Our approach allows a priori specification of the maximally allowed distortion, and it automatically finds a sufficient representative subset of the data to act as codebook vectors (or cluster centres). It does not find the minimal number of such vectors, which would amount to a combinatorial problem; however, we find a ’good’ quantization through linear programming.

Cite this Paper

BibTeX
@InProceedings{pmlr-vR3-tipping01a, title = {A Kernel Approach for Vector Quantization with Guaranteed Distortion Bounds}, author = {Tipping, Michael E. and Sch{\"{o}}lkopf, Bernhard}, booktitle = {Proceedings of the Eighth International Workshop on Artificial Intelligence and Statistics}, pages = {298--303}, year = {2001}, editor = {Richardson, Thomas S. and Jaakkola, Tommi S.}, volume = {R3}, series = {Proceedings of Machine Learning Research}, month = {04--07 Jan}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/r3/tipping01a/tipping01a.pdf}, url = {https://proceedings.mlr.press/r3/tipping01a.html}, abstract = {We propose a kernel method for vector quantization and clustering. Our approach allows a priori specification of the maximally allowed distortion, and it automatically finds a sufficient representative subset of the data to act as codebook vectors (or cluster centres). It does not find the minimal number of such vectors, which would amount to a combinatorial problem; however, we find a ’good’ quantization through linear programming.}, note = {Reissued by PMLR on 31 March 2021.} }
Endnote
%0 Conference Paper %T A Kernel Approach for Vector Quantization with Guaranteed Distortion Bounds %A Michael E. Tipping %A Bernhard Schölkopf %B Proceedings of the Eighth International Workshop on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2001 %E Thomas S. Richardson %E Tommi S. Jaakkola %F pmlr-vR3-tipping01a %I PMLR %P 298--303 %U https://proceedings.mlr.press/r3/tipping01a.html %V R3 %X We propose a kernel method for vector quantization and clustering. Our approach allows a priori specification of the maximally allowed distortion, and it automatically finds a sufficient representative subset of the data to act as codebook vectors (or cluster centres). It does not find the minimal number of such vectors, which would amount to a combinatorial problem; however, we find a ’good’ quantization through linear programming. %Z Reissued by PMLR on 31 March 2021.
APA
Tipping, M.E. & Schölkopf, B.. (2001). A Kernel Approach for Vector Quantization with Guaranteed Distortion Bounds. Proceedings of the Eighth International Workshop on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research R3:298-303 Available from https://proceedings.mlr.press/r3/tipping01a.html. Reissued by PMLR on 31 March 2021.

Related Material