第17卷第1期 Volume 17 Number 2020年1月 铁道科学与工程学报 Journal of Railway Science and Engineering January 2020 doi:10.19713/.cni.43-1423/u.20190248 回回 基于大数据的铁路工程投标企业 异常行为预警研究 乔柱,刘伊生,茹建青2 (1.北京交通大学经济管理学院,北京100044 2.国家铁路局工程质量监督中心,北京100891) 摘要:基于铁路工程电子招投标积累的海量数据,识别预警具有异常行为的铁路工程投标企业。运用社团结构检测分析识 别具有抱团行为的投标企业;再通过函数拟合预测投标企业中标次数,根据实际值与预测值的标准残差所落置信区间范围 分析识别中标次数异常的投标企业;最后综合分析2种方法的结果,构建铁路工程投标企业异常行为分级预警模型,识别具 有串围标嫌疑的企业并分级预警。研究结果表明基于铁路工程电子招投标大数据,综合社团结构检测和函数拟合分析两和 方法,可以有效识别预警铁路工程投标企业的异常行为。 关键词:大数据;投标企业;异常行为;预警 中图分类号:C931文献标志码:A文章编号:1672-7029(20)01-0250-08 Research on early warning of abnormal behavior of railway engineering bidding enterprises based on big data QIAO Zhu', LIU Yisheng', RU Jianqing (1. School of Economics and Management, Beijing Jiaotong University. Beijing 100044. China Abstract: Based on the accumulated data of railway engineering electronic bidding, the railway engineering bidding enterprises with abnormal behavior was identified and early warned. Firstly, this paper used the community structure detection to analyze and identify the bidding enterprises with group behavior. Secondly, by redicting the successful bid number of bidding enterprises by function fitting, and calculating the location of the standard residuals of actual value and predicted value for the confidence interval range, the bidding enterprise with the abnormal number of successful bids was analyzed and identified. Finally, considering the results of the two methods, a g an early classification warning model for the abnormal behavior of railway engineering bidding enterprises, e nder suspicion of together-conspired bidding and contacting bid was identified and early classification warned. The research shows that the railway project electronic bidding big data can effectively identify and early warn the railway engineering bidding enterprises with abnormal behavior, comprehensively using the two methods which includes community structure detection and function fitting analysis. Key words: big data; bidding enterprise; abnormal behavior; early warning 收稿日期:2019-04-01 基金项目:国家自然科学基金资助项目(784102 通信作者:乔柱(1991-),男,河南南阳人,博士研究生,从铁路工程招投标研究:E-mai:151063687212.cm