On the Combination of Logistic Regression and Local Probability Estimates

Melanie Osl, C. Baumgartner, B. Tilg, Stephan Dreiseitl

Publikation: Beitrag in Buch/Bericht/TagungsbandKonferenzbeitragBegutachtung

2 Zitate (Scopus)

Abstract

Classifiers based on parametric or non-parametric learning methods have different advantages and disadvantages. To take advantage of the strengths of both methods, we propose an algorithm that combines a parametric model (logistic regression) with a non-parametric classification method (k-nearest neighbors). This combination is based on a measure of appropriateness that uses a heuristic to decide which of the two components should contribute more to the final classification output. We measure the performance of this combination method on two data sets (one from medical informatics, and one consisting of simulated data) in terms of areas under the ROC curves (AUCs). We are able to demonstrate that our method of combining classifiers exceeds the performance of both individual classifiers taken separately.

OriginalspracheEnglisch
TitelProceedings - 3rd International Conference on Broadband Communications, Informatics and Biomedical Applications, BroadCom 2008
Seiten124-128
Seitenumfang5
DOIs
PublikationsstatusVeröffentlicht - 2008
VeranstaltungBroadcom 2008 - International Conference on Broadband Communications, Information Technology and Biomedical Applications - Pretoria, Südafrika
Dauer: 3 Nov 20086 Nov 2008

Publikationsreihe

NameProceedings - 3rd International Conference on Broadband Communications, Informatics and Biomedical Applications, BroadCom 2008

Konferenz

KonferenzBroadcom 2008 - International Conference on Broadband Communications, Information Technology and Biomedical Applications
Land/GebietSüdafrika
OrtPretoria
Zeitraum03.11.200806.11.2008

Fingerprint

Untersuchen Sie die Forschungsthemen von „On the Combination of Logistic Regression and Local Probability Estimates“. Zusammen bilden sie einen einzigartigen Fingerprint.

Zitieren