Computation on Attribute Importance of Classification Based on Cloud Model

Different condition attributes in classification of data sets often have different importance in relation to the decision attributes. Not all condition attributes are necessary in constructing classification rules, excluding some attributes will not affect the classification results. Thus attribute reduction is an important pre-treatment process in classification mining. It can delete redundant attributes on the basis of maintaining steady classification abilities of decision system. This paper presents a method to calculate data sets similarity based on cloud model.