Research On China's GDP Based On Cluster Analysis

As the core indicator of national economic accounting,gross domestic product(GDP)has always been a hot issue in various disciplines.Based on the clustering analysis algorithm in multivariate statistical analysis,this paper has done a lot of research on several major GDP data indicators in various provinces in China,and made empirical analysis.The paper first introduces various algorithms of cluster analysis,and focuses on two systems clustering methods(intermediate distance method and class average method)and related theories of K-means clustering algorithm,and introduces the Dunn index as an index to measure the clustering effect,so as to compare the three clustering algorithms mentioned above.Then,this paper selects the per capita GDP data of 31 provinces in China for nearly 20 years,and uses the three clustering analysis algorithms to classify the per capita GDP data of all provinces first.After comparing the Dunn index,the K-means algorithm has the most clustering effect.Therefore,this algorithm is applied to the subsequent content of this article as the most important algorithm used in this paper.In addition,this paper also makes a cluster analysis on the total GDP data of each province,and compares this classification result with the classification result of per capita GDP data.It is found that there are certain differences between the two results:in terms of total GDP,the east coastal provinces of the region have obvious advantages and occupy the optimal or medium category.In terms of per capita GDP,the municipality largely avoids the burden of population,so they naturally occupy the optimal category.However,most of the provinces in the western region are in a backward category,and the situation of uncoordinated regional development needs to be taken seriously.Next,the paper divides the total GDP data into the added value of the three major industries,and continues to use the K-means algorithm to conduct a clustered research and empirical analysis of the industrial value-added data of the provinces in the past 20 years.The results show that there is a certain difference in the echelons of the added value of the three major industries in each province.However,in general,the added value of the three major industries in the eastern provinces and the central provinces are at a relatively leading level,while in the western provinces,the added value of the three major industries are relatively backward.From the perspective of industries,the differences in regional development are also obvious.In addition,this paper also calculates and summarizes the GDP share of the three major industries in the past five years,and compares the differences in GDP structure between different provinces.The results show that the proportion of tertiary industry in Beijing and Shanghai is obviously superior to other provinces,and the industrial structure is the most advantageous.The proportion of the secondary industry and the tertiary industry in most other provinces is not much different,and it is much higher than the proportion of the primary industry.It reflects that the secondary industry and the tertiary industry are the main support points of GDP growth at present,and the industrial structure is becoming more and more reasonable.Finally,this paper discusses some inequalities about g-expectation,including Holder's inequality,Minkowski's inequality,and some of their inferences.On the innovation of this paper,firstly,three different clustering analysis algorithms are selected to cluster the per capita GDP data of each province,and the results arc compared to select the best clustering algorithm.In addition,this paper fully considers a number of indicators related to GDP data.On the basis of the data of per capita GDP and gross GDP of each province,this paper further makes cluster analysis of the value added of the three major industries and their proportion of GDP in each province,and draws more specific conclusions,which have stronger guiding significance.