Quality-first pattern-based clustering approach with fuzzed thresholds

Abstract

Pattern-based clustering has broad applications such as in DNA micro array data analysis, customer segmentation, e-business data analysis, etc. However, pattern-based clustering often returns a large number of highly-overlapping clusters, which makes it hard for users to identify interesting patterns from the mining results. One of the most important factors which cause highly-overlapping is the error thresholds defined. This paper makes error analysis in pattern-based clustering. With a fuzzed thresholds strategy to deduce overlapping which caused by error range, a quality-first clustering algorithm is proposed to mining pclusters of higher quality. Both theoretical and experimental analysis shows that our method is effective and efficient. © 2011 Springer-Verlag.