Research on Shenzhen Metro Station Clustering Based on Multi-Source Big Data and Unsupervised Clustering Algorithm

Based on the multi-source big data and unsupervised clustering Algorithm, the characteristics of Shenzhen subway stations are discussed to conclude the development of the Shenzhen subway. First, the number of passengers entering and leaving the Shenzhen Metro in the morning peak, 8 types of Point of Interest (POI) at each station, and the transaction amount are selected as cluster items for cluster analysis and processing. Secondly, the characteristics of each category are analyzed according to the clustering results. The results suggest that: (1) According to the clustering index, the subway stations are divided into residence-oriented, employment-oriented, jobs-housing dislocation, dislocation partial residential, dislocation partial employment, mixed type, comprehensive type, transportation hub type, and others. There are nine different types; (2) POI data can well describe the site's points of interest characteristics within 1 km; (3) Make recommendations for future passenger flow and management of specific sites.