基于谱聚类算法的新疆增长极层级划分方法  

A Method of Hierarchy Division of Xinjiang Growth Pole Based on Spectral Clustering Algorithm

作　　者：刘凯[1] LIU Kai(School of Statistics and Data Science,Xinjiang University of Finance and Economics,Urumqi Xinjiang 830012)

机构地区：[1]新疆财经大学统计与数据科学学院,新疆乌鲁木齐830012

出　　处：《数字技术与应用》2021年第3期102-104,共3页Digital Technology & Application

摘　　要：城市化是新疆地区当前与未来发展的重大课题,用何种发展思路与全局发展模式显得尤为重要。文章在参考与理解部分学者对层级增长极理论的研究基础上,探索性的采用机器学习中的无监督学习模型——谱聚类算法进行统计建模,从定量的角度设计一种对区域增长极网络层级划分方法;并通过对新疆部分县市的实证分析,得到新疆层级增长极网络的结构。Urbanization is a major issue for the current and future development of Xinjiang.What kind of development thinking and overall development model are particularly important.Based on the reference and understanding of some scholars’research on the theory of hierarchical growth poles,this article exploratively uses the spectral clustering algorithm in the unsupervised learning model for statistical modeling,and designs a quantitative division of the regional growth pole network hierarchy method;and through empirical analysis of some counties and cities in Xinjiang,the structure of Xinjiang's hierarchical growth pole network is obtained.

关 键 词：谱聚类算法 层级增长极网络 新疆城市化 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] F061.5[自动化与计算机技术—控制科学与工程]