The Dynamic KNN Clustering of Undergraduate Consumption with Gini Coefficient: A Case of Zhejiang

Combining the nonparametric KNN-kernel density-based clustering method (Tranh, 2005) with Gini coefficient formula (Zuguang Hu, 2004), we investigate how Gini coefficients affect the classification among 31 areas in Zhejiang province based on the undergraduate consumption data. The results show that there is 10 percent classification difference between the classification with Gini coefficient and the one not, then we get the desired classification through adjusting the Gini coefficient suitably and dynamically.