DocumentCode :
437475
Title :
Stability of the classifier based on modification on Schweizer and Sklars equations
Author :
Luukka, P. ; Sampo, J.
Author_Institution :
Lab. of Appl. Math., Lappeenranta Univ. of Technol.
Volume :
1
fYear :
2004
fDate :
1-3 Dec. 2004
Firstpage :
207
Lastpage :
212
Abstract :
In this article we have applied Schweizer & Sklars implications with extension to generalized mean to classification task. We show that classification results are not so sensitive to p values with Schweizer & Sklars measures, which indicates generalized form of equations. Investigation for correct mean values is carried out. In this article we have also tested stability of the classifier. Two different tests for stability are made: in one test stability was checked respect to weight parameters and other test was carried out for ideal vectors
Keywords :
learning (artificial intelligence); pattern classification; statistical testing; Schweizer & Sklars equation; machine learning; pattern classification; stability; statistical testing; Computer errors; Diseases; Equations; Humans; Laboratories; Manufacturing processes; Mathematics; Pipelines; Stability; Testing;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Cybernetics and Intelligent Systems, 2004 IEEE Conference on
Conference_Location :
Singapore
Print_ISBN :
0-7803-8643-4
Type :
conf
DOI :
10.1109/ICCIS.2004.1460413
Filename :
1460413
Link To Document :
بازگشت