Analisis Pemilihan Pemboleh Ubah dalam Proses Pembentukan Sistem Geodemografi

Selection Analysis of Variables to Produces Geodemography System

Abstract

Geodemography classification system (SPG) is refers to the classification of the population according to the location or where they live. This statement refers to the tendency of people who are more comfortable to mix, forming communities and residents who live in the area have in common with them. The concept of this study are similar to the first law of geography presented by Tobler in 1970, that all things are related to other things, but the things that are closer have a stronger relationship than with more distant things. Data population and housing census is the most comprehensive database of the collected continuously from 1870 until 2010. The analysis of data selection is done because many of the variables that will cause the repetition of the same information. Through this study, analysis of data selection such as principal component analysis (PCA), correlation analysis and analysis of variance carried out to ensure that the variables that really should be only used in forming the classification system geodemografi in Perak. 179 variables in a population and housing census data Perak in 2000 to develop a classification system geodemography (SPG) and from the analysis, only 69 potential variables selected to run the cluster analysis. Reducing the number of variables in the analysis variable selection will simplify the process of cluster analysis was performed