Title :
A Fuzzy Genetic Clustering Technique Using a New Symmetry Based Distance for Automatic Evolution of Clusters
Author :
Saha, Sriparna ; Bandyopadhyay, Sanghamitra
Author_Institution :
Machine Intelligence Unit, Indian Stat. Inst., Calcutta
Abstract :
In this paper a fuzzy point symmetry based genetic clustering technique (fuzzy-VGAPS) is proposed which can determine the number of clusters present in a data set as well as a good fuzzy partitioning of the data. A new fuzzy cluster validity index, FSym-index, which is based on the newly developed point symmetry based distance is also proposed here. FSym-index provides a measure of goodness of clustering on different fuzzy partitions of a data set. Maximum value of FSym-index corresponds to the proper clustering present in a data set. The flexibility of fuzzy-VGAPS is utilized in conjunction with the fuzzy FSym-index to determine the number of clusters present in a data set as well as a good fuzzy partition of the data. The results of the fuzzy VGAPS are compared with those obtained by fuzzy version of variable string length genetic clustering technique (fuzzy-VGA) optimizing XB-index. The effectiveness of the fuzzy-VGAPS is demonstrated on four artificial data sets and two real-life data sets
Keywords :
data mining; fuzzy set theory; genetic algorithms; pattern clustering; data clustering; data fuzzy partitioning; fuzzy FSym-index; fuzzy cluster validity index; fuzzy genetic clustering technique; fuzzy point symmetry based distance; variable string length genetic clustering technique; Biological cells; Clustering algorithms; Data mining; Data structures; Euclidean distance; Fuzzy sets; Genetic algorithms; Machine intelligence; Particle measurements; Shape; Cluster Validity Index; Clustering; Genetic Algorithm; Kd-tree; Point Symmetry; Variable String Length;
Conference_Titel :
Computing: Theory and Applications, 2007. ICCTA '07. International Conference on
Conference_Location :
Kolkata
Print_ISBN :
0-7695-2770-1
DOI :
10.1109/ICCTA.2007.5
