摘要
本文将数据挖掘中的决策树分类方法运用到工程项目评标数据分析,从200多个天津市工程项目招投标打分数据中,随机抽取15个招投标项目中的67个承包商的评标专家打分数据进行分析,得到中标承包商技术和商务评分分界点,进而得到工程项目潜在风险的预警阈值,然后借助因子分析辨识出风险来源并进行预警。
In this paper, a model based on decision trees in data mining was applied to the analysis of the tender evaluation data of civil projects in Tianjin, 67 contractors among 15 projects were selected randomly from 200 civil projects. The critical values of the winners' indices, such as technique and business were obtained by CART algorithm in the model, then the potential risks of the civil project could be identified by Factor Analysis and some advice was given.
出处
《数理统计与管理》
CSSCI
北大核心
2010年第1期122-128,共7页
Journal of Applied Statistics and Management
基金
天津市水利局科研项目资助(项目编号:KY2007-09)
关键词
决策树分类
CART算法
评标
风险预警
decision tree classification, CART algorithm, tender evaluation, risk pre-warning