Intelligent clustering of Spatial data for spatial classification using Data Mining

Shafeeq, Ahamed B M and Hegde, Govardhan (2011) Intelligent clustering of Spatial data for spatial classification using Data Mining. International Journal of Earth Sciences and Engineering, 4 (11). pp. 397-403. ISSN 0974-5904

[img] PDF
shafeeq1.pdf - Published Version
Restricted to Registered users only

Download (187kB) | Request a copy


Clustering large spatial databases is an important problem which tries to find the densely populated regions in the feature space to be used in data mining. The aim of this paper is to study the regional economic difference with the spatial data mining theories and classify the regions intelligently for the sustainable development of the province. This information can be used by the administrative department for launching the various plans for the development of the province. In this paper, we take the per capita GDP as index variable, and take the state administration as basic analysis unit and clustering the states based on per capita GDP. Based on the ESDA methods(global and local spatial autocorrelation) of spatial data mining theory, including Moran I index, Moran scatter plot and LISA, we research and analyze the economy of spatial distribution of India in 2009 from the spatial interactive angle. The results show that economy of Indian states has a strong spatial correlation and there also exist spatial heterogeneity problems between states. Clustering is an important task in spatial data mining and spatial analysis. A family of bioinspired algorithms, well-known as Swarm Intelligence (SI) has recently emerged that meets these requirements and has successfully been applied to a number of real world clustering problems. A very important analysis on state’s economy data would be clustering to identify state groups with the attributes of interest which have similar properties so that the central administration can plan for the sustainable development of the province. In contrast to the prevalent research efforts of developing new algorithms, there has been a lack of effort to re-use existing algorithms for varying domain and tasks. One such intelligent clustering algorithm is the Automatic Clustering with Particle Swarm Optimization (MEPSO) algorithm which was recently proposed as a very good candidate to solve such problems.

Item Type: Article
Uncontrolled Keywords: Spatial data mining, Spatial autocorrelation, ESDA, Clustering
Subjects: Engineering > MIT Manipal > Computer Science and Engineering
Depositing User: MIT Library
Date Deposited: 04 Sep 2014 05:45
Last Modified: 04 Sep 2014 05:45

Actions (login required)

View Item View Item