Static field approach for pattern classification

Authors: Ruta, D. and Gabrys, B.

Journal: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Volume: 2311

Pages: 232-246

eISSN: 1611-3349

ISBN: 9783540434818

ISSN: 0302-9743

DOI: 10.1007/3-540-46019-5_18

Abstract:

Recent findings in pattern recognition show that dramatic improvementfof the recognition rate can be obtained by application of fusion systemsfutilizing many different and diverse classifiers for the same task. Apart from afgood individual performance of individual classifiers the most important factorfis the useful diversity they exhibit. In this work we present an example of afnovel non-parametric classifier design, which shows a substantial level of diversityfwith respect to other commonly used classifiers. In our approach inspirationffor the new classification method has been found in the physical world. Namelyfwe considered training data as particles in the input space and exploited the conceptfof a static field acting upon the samples. Specifically, every single datafpoint used for training was a source of a central field, curving the geometry offthe input space. The classification process is presented as a translocation in thefinput space along the local gradient of the field potential generated by the trainingfdata. The label of a training sample to which it converged during the translocationfdetermines the eventual class label of the new data point. Based on selectedfsimple fields found in nature, we show extensive examples and visual interpretationsfof the presented classification method. The practical applicabilityfof the new model is examined and tested using well-known real and artificialfdatasets.

Source: Scopus

Static field approach for pattern classification

Authors: Ruta, D. and Gabrys, B.

Journal: SOFT-WARE 2002: COMPUTING IN AN IMPERFECT WORLD

Volume: 2311

Pages: 232-246

ISSN: 0302-9743

Source: Web of Science (Lite)

Static Field Approach for Pattern Classification

Authors: Ruta, D. and Gabrys, B.

Editors: Bustard, D., Liu, W. and Sterritt, R.

Volume: 2311/2002

Pages: 232-246

Publisher: Springer

Place of Publication: Berlin

ISBN: 978-3-540-43481-8

DOI: 10.1007/3-540-46019-5

Abstract:

Recent findings in pattern recognition show that dramatic improvement of the recognition rate can be obtained by application of fusion systems utilizing many different and diverse classifiers for the same task. Apart from a good individual performance of individual classifiers the most important factor is the useful diversity they exhibit. In this work we present an example of a novel non-parametric classifier design, which shows a substantial level of diversity with respect to other commonly used classifiers. In our approach inspiration for the new classification method has been found in the physical world. Namely we considered training data as particles in the input space and exploited the concept of a static field acting upon the samples. Specifically, every single data point used for training was a source of a central field, curving the geometry of the input space. The classification process is presented as a translocation in the input space along the local gradient of the field potential generated by the training data. The label of a training sample to which it converged during the translocation determines the eventual class label of the new data point. Based on selected simple fields found in nature, we show extensive examples and visual interpretations of the presented classification method. The practical applicability of the new model is examined and tested using well-known real and artificial datasets.

Source: Manual

Preferred by: Dymitr Ruta

Static Field Approach for Pattern Classification.

Authors: Ruta, D. and Gabrys, B.

Editors: Bustard, D.W., Liu, W. and Sterritt, R.

Journal: Soft-Ware

Volume: 2311

Pages: 232-246

Publisher: Springer

https://doi.org/10.1007/3-540-46019-5

Source: DBLP