摘要
通过对公交刷卡数据进行分析,提取特征向量,用斯皮尔曼相关系数分析向量间的相关性,数据优化后采用大数据技术实现对K-means聚类算法的深入挖掘,最终得出K=6,即将用户分为6类最为合理,通过对公交用户出行数据的详细解读,得到每类用户的出行画像,使行业管理者和决策者能够更加清晰、准确地了解用户特征,并制定出有针对性的决策.
By analyzing the bus farecard transaction data,the feature vector is extracted,and the correlation between vectors is analyzed by using the Spearman correlation coefficient.After the data is optimized,the big data analytics are used to realize the deep mining of K-means clustering algorithm,and finally K is calculated to be 6,which means that users can be divided into 6 categories.Through the detailed interpretation of the travel data of bus users,the travel portraits of each type of users can be obtained,so that industry managers and decision makers can understand user characteristics more clearly and accurately,and work out targeted decision making.
作者
范桂莲
马跃
FAN Guilian;MA Yue(Wuhan Institute of Transportation Science,Wuhan 430014,China;The Beijing Tong Tu Soft,LLC.,Beijing 100080,China)
出处
《交通工程》
2023年第3期71-76,共6页
Journal of Transportation Engineering