Title :
Fuzzy cluster analysis with missing values
Author :
Timm, H. ; Kruse, R.
Author_Institution :
Fac. of Comput. Sci., Magdeburg Univ. of Technol., Germany
Abstract :
The paper shows how methods of fuzzy cluster analysis can be extended in such a way that they are able to process datasets with missing values. We study several imputation methods and propose a new approach for reducing the influence of imputations
Keywords :
data analysis; fuzzy set theory; pattern recognition; data analysis; datasets; fuzzy cluster analysis; fuzzy clustering algorithms; imputation methods; missing values; Clustering algorithms; Computer science; Data analysis; Data preprocessing; Ellipsoids; Noise robustness; Noise shaping; Shape;
Conference_Titel :
Fuzzy Information Processing Society - NAFIPS, 1998 Conference of the North American
Conference_Location :
Pensacola Beach, FL
Print_ISBN :
0-7803-4453-7
DOI :
10.1109/NAFIPS.1998.715573
