An analysis of regional product by Three Strata of Industry based on K-means method

This study applies cluster analysis K-means method to analyze the data of regional product by Three Strata of Industry in China. Firstly, we give a review of K-means method, according to self-organizing feature maps we determine the number of clusters, and then employ the K-means method to find the clusters. From the cluster analysis we get four clusters and conclude that the membership in Cluster 1 are all municipalities with the highest value of Per Capita GRP and the highest percentage composition of Tertiary Industry. It shows that the economy of this cluster is developed. Cluster 2 with the lowest composition of Tertiary Industry should enhance the development of Tertiary Industry. The Secondary Industry of Cluster 3 is the most developed. The most provinces in China concentrate in Cluster 4 with the lowest Per Capita GRP need to enhance both Secondary Industry and Tertiary Industry.