A Linear Regression Model for Spatial Data Mining: An Experimental Approach

Abstract: A GIS data may be collection of spatial and non spatial data types. Feature extraction of spatial data types from a huge data is called spatial data mining. Spatial data mining has accepted as very new and emerging technology for development of system which are applicable directly or indirectly in various field of human needs as e-marketing, cluster analysis of population density , cost estimation of land with forest clustering, geographic trend detection etc. This paper is based on spatial analysis of a city locations which are linearly auto correlated with each other. A real data set has used in this paper and a linear auto correlation is shown between them. Two major attributes as longitude and latitude are selected for experimental purpose. A model is also designed for experimental setup. Experimental setup is completed in PYTHON with graphical touch. Results are very specific and supporting with author's key objectives.

Keywords: Linear regression, Auto correlation, PYTHON, Spatial Data mining