The Analysis Of Data Based On The Hierarchical Clustering

With the rapid development of computer, computer technology has been used for electronic commerce, business management, scientific research and government office and other fields. Especially in the Internet technology becomes more mature, there are a lot of even large amounts of data on the Internet, people in order to obtain useful information and knowledge, they must solve the contradiction between excessive information and poor knowledge, so the birth of a new subject--data mining. Data mining as a new discipline, has its own procedures and methods.The purpose of Data mining is produced for reference, therefore the method of data mining,which is good or bad, mainly depends on the actual application effect Data mining is used very widely, especially in the economic field.Data mining has become the focus of today’s social studiesThe basic methods of data mining include classification, clustering, regression analysis, association rule mining, feature, web page, change and deviation analysis.This paper is mainly to study the clustering algorithm based on data mining, firstly describes the current status of data mining, and introduces the concept, basic methods and applications. Then, introduces several common clustering algorithm. The combined application of hierarchical clustering algorithm and analytic hierarchy process and a hierarchical clustering based on cloud model and through the analysis of the experiments to verify the proposed. Last, through the data of UCI experiment analysis and comparison, obtain the feasibility and effectiveness of the algorithm in this paper. The main research contents of this paper are as follows:Firstly, a method of improved Eigenvalue method(EM) based on Analytic hierarchy process(AHP) is presented in the paper.Given the combined application of hierarchical clustering and analytic hierarchy process. In order to make up for the deficiency of single application of AHP.Secondly, in order to apply the hierarchical clustering algorithm is applied to the fuzzy set theory and the analysis of properties of the large amounts of high-dimensional data,after clustering. A method which a hierarchical clustering algorithm based on cloud model is presented in the paper. The cloud model combined the randomness and fuzziness of natural law, it’s Study belongs to the fuzzy clustering. It will be a hierarchical clustering algorithm is successfully applied to the fuzzy set theory in the paper, and gives the method to judge the degree of membership. The experimental results show that, Hierarchical clustering of cloud model has good clustering effect, with the increasing of data, its complexity is much lower than the original hierarchical clustering algorithm, This can make big data clustering...