Fuzzy Robust Statistics for Application to the Fuzzy c-Means Clustering Algorithm

Abstract
This report introduces two robust statistics -- the fuzzy median and the fuzzy median absolute deviation from the median -- that have been developed for use with fuzzy data sets. The two statistics were applied to the fuzzy c- Means algorithm, a powerful clustering algorithm that normally employs linear statistics. The modified algorithm showed improved performance, being able to cluster data sets generated by heavy-tailed distributions like the Cauchy and Slash distributions.

This publication has 0 references indexed in Scilit: