Stationarity of Matrix Relevance LVQ

Biehl M, Hammer B, Schleif F-M, Schneider P, Villmann T (2015)
In: 2015 International Joint Conference on Neural Networks (IJCNN). IEEE.

Download
Es wurde kein Volltext hochgeladen. Nur Publikationsnachweis!
Konferenzbeitrag | Veröffentlicht | Englisch
Autor
; ; ; ;
Abstract / Bemerkung
We present a theoretical analysis of Learning Vector Quantization (LVQ) with adaptive distance measures. Specifically, we consider generalized Euclidean distances which are parameterized in terms of a quadratic matrix of adaptive relevance parameters. Winner-takes-all prescriptions based on the heuristic LVQl are in the center of our interest. We derive and study stationarity conditions and show, among other results, that stationary prototypes can be written as linear combinations of the training data apart from irrelevant contributions in the null-space of the relevance matrix. The investigation of the metrics updates reveals that relevance matrices become singular with only one or very few non-zero eigenvalues. Implications of this property are discussed and, furthermore, the effect of preventing singularity by introducing an appropriate penalty term is studied. Theoretical findings are confirmed in terms of illustrative example data sets.
Erscheinungsjahr
Titel des Konferenzbandes
2015 International Joint Conference on Neural Networks (IJCNN)
Konferenz
2015 International Joint Conference on Neural Networks (IJCNN)
Konferenzort
Killarney, Ireland
Konferenzdatum
2015-07-12 – 2015-07-17
PUB-ID

Zitieren

Biehl M, Hammer B, Schleif F-M, Schneider P, Villmann T. Stationarity of Matrix Relevance LVQ. In: 2015 International Joint Conference on Neural Networks (IJCNN). IEEE; 2015.
Biehl, M., Hammer, B., Schleif, F. - M., Schneider, P., & Villmann, T. (2015). Stationarity of Matrix Relevance LVQ. 2015 International Joint Conference on Neural Networks (IJCNN) IEEE. doi:10.1109/ijcnn.2015.7280441
Biehl, M., Hammer, B., Schleif, F. - M., Schneider, P., and Villmann, T. (2015). “Stationarity of Matrix Relevance LVQ” in 2015 International Joint Conference on Neural Networks (IJCNN) (IEEE).
Biehl, M., et al., 2015. Stationarity of Matrix Relevance LVQ. In 2015 International Joint Conference on Neural Networks (IJCNN). IEEE.
M. Biehl, et al., “Stationarity of Matrix Relevance LVQ”, 2015 International Joint Conference on Neural Networks (IJCNN), IEEE, 2015.
Biehl, M., Hammer, B., Schleif, F.-M., Schneider, P., Villmann, T.: Stationarity of Matrix Relevance LVQ. 2015 International Joint Conference on Neural Networks (IJCNN). IEEE (2015).
Biehl, Michael, Hammer, Barbara, Schleif, Frank-Michael, Schneider, Petra, and Villmann, Thomas. “Stationarity of Matrix Relevance LVQ”. 2015 International Joint Conference on Neural Networks (IJCNN). IEEE, 2015.

Export

Markieren/ Markierung löschen
Markierte Publikationen

Open Data PUB

Suchen in

Google Scholar
ISBN Suche