Fuzzy projective clustering in high dimension data using decrement size of data
S. Mohaddese Seyednejad, Hamid Musavi, S. Mohaddese Seyednejad, Tooraj Darabi · 2011
Today, data clustering problems became an important challenge in Data Mining domain. A kind of clustering is projective clustering. Since a lot of researches has done in this article but each of previous algorithms had some defects that we will be indicate in this paper. We propose a new algorithm based on fuzzy sets and at first using this approach detect and eliminate unimportant properties for all clusters. Then we remove outliers, finally we use weighted fuzzy c-mean algorithm according to offered formula for fuzzy calculations. Experimental results show that our approach has more performance and accuracy than similar algorithms.