Research of the Dimension Combination Strategy Model

Abstract

This study quantifies the usage and evaluation of Data Reduction. Within Data Reduction, there are three different measurement of methods: association measurement, discrimination measurement, and information measurement. Through analysis of the importance of each measurement stage, we generated sequences of forward generation to select the best combination of Data Reduction. The purpose of the sequences of forward generation is to increase efficiency and accuracy from the selected combination of Data Reduction. Based on the method of generating our model, we want only a single field to appear, in order to measure the amount of information based on the most suitable model law for the three measurement methods. The purpose of this model is to allow users of data mining to explore the selected field, in addition to the single characteristic attribute field as a reference, but also according to different dimensions of the resulting combination of all the chaos of the target attributes and how they affect the relationship, so that users can analyze and use the field to solve the most troublesome mining field dimension selections.