Visible to the public Fuzzy clustering of incomplete data based on missing attribute interval size

TitleFuzzy clustering of incomplete data based on missing attribute interval size
Publication TypeConference Paper
Year of Publication2015
AuthorsZhang, L., Li, B., Zhang, L., Li, D.
Conference Name2015 IEEE 9th International Conference on Anti-counterfeiting, Security, and Identification (ASID)
Date Published Sept. 2015
ISBN Number978-1-4673-7140-7
KeywordsAlgorithm design and analysis, Breast, Clustering algorithms, data handling, data structures, Fuzzy C-Means, fuzzy c-means algorithm, fuzzy clustering, fuzzy set theory, Incomplete Data, incomplete data clustering, interval data set, interval median, Interval size, Iris, missing attribute interval size, nearest neighbor rule, pattern clustering, Prototypes, pubcrawl170107, similar object cluster identification, Standards, TICI data set

Fuzzy c-means algorithm is used to identity clusters of similar objects within a data set, while it is not directly applied to incomplete data. In this paper, we proposed a novel fuzzy c-means algorithm based on missing attribute interval size for the clustering of incomplete data. In the new algorithm, incomplete data set was transformed to interval data set according to the nearest neighbor rule. The missing attribute value was replaced by the corresponding interval median and the interval size was set as the additional property for the incomplete data to control the effect of interval size in clustering. Experiments on standard UCI data set show that our approach outperforms other clustering methods for incomplete data.

Citation Keyzhang_fuzzy_2015