Methods Research Of Intelligent Statistic Analysis Of Enterprise Information In The Electronic Business Environment

This paper is sponsored by Chongqing science and technology soft project (No: 7117). And most content has appeared on The Second International Conference on Machine Learning and Cybertics.This paper has been searched whole-length by EIElectronic Business has been developing so fast in the 21st century. With its fast development, the enterprise information becomes chaos, abundant,What's more, the enterprise information has a lot of deficiencies such as fast transmission speed, verboseness and indefiniteness. Therefore statistic analysis of enterprise information to support decision-making is a difficult problem which enterprise must face.How to deal with the information in the Electronic Business environment has been a topic of general interest at home and abroad. But the existing theories and methods are to some extent limited when they are put into practical use. So how to carry out the research of method of enterprise information'intelligent statistic analysis in the light of its characteristics of enterprise information in the Electronic Business environment will be a big question for study or discussion before us.With careful considerations of the characteristics of enterprise information in the Electronic Business environment and with Rough set theory, Statistical learning theory,support vector machine as the theoretical methods, the author of this paper carries out the research of intelligent statistic analysis of enterprise information mainly from 3 aspects. The first: data preprocess of enterprise information in the Electronic Business environment. We deal with the problems of lost data and abnormal data in original information by combining the statistic method and genetic algorithm. The second: the reduction of enterprise information and how to distinguish the useful information in the Electronic Business environment. We make use of the information entropy in the Rough set and evaluate the degree of importance of information so as to reduce information. Then we pick up the useful information by adopting improved least square support vector machine from reduction information. The third: the foundation of intelligent statistic analysis model of enterprise information on the basis of the information which has been reduced and distinguished. The model is based on Support Vector Machine and realize Structure Risk Minimization principle. Being put into practical use, the model proves to be applicable.The main innovation of this paper is as follow: 1.Give the method of Least Square Support Vector Machine to pick-up valuable information and complex calculation is overcame; 2. Bring forword the method of muti-factors dynamic forecasting. 3. Introduce the support vector machine to enterprise information analysis in the Electronic Business environment. Put forward an intelligent statistic analysis method.