Fuzzy-GSO Algorithm for Mining of Irregularly Shaped Spatial Clusters

Abstract

Spatial clustering is one of the vital methods for spatial data mining and spatial data analysis. Spatial clustering aims to partition spatial data into a series of meaningful subclasses, called spatial clusters, such that spatial objects in the same cluster are similar to each other, and are dissimilar to those in different clusters. Literature presents several algorithms to spatial data clustering using density computation, spatial proximity and attributes similarity. Recently, optimization algorithm called ant colony optimization was used for spatial data clustering utilizing spatial scan statistic (SaTScan) which has become one of the most popular methods for detecting and evaluating spatial clusters. But, SaTScan is well suitable to identify only the circular or elliptical clusters and it not much effective for the detection of irregularly shaped clusters. So, in order to tackle this challenge, an algorithm will be developed to be suitable for irregular shaped data using recent optimization algorithm, GSO. In this paper a Fuzz-GSO algorithm is developed to cluster the spatial objects. The GSO algorithm performs the operations such as producer, scrounger and ranger. The fuzzy operator is also incorporated into the GSO algorithm to enhance the performance of clustering.