摘要
以县域中小企业为样本数据,应用Logistic回归模型、决策树模型、k近邻分类模型比较分析中小企业生存时间与其注册资本、产业类别、企业类别和企业类型之间的关系,并从政府和企业的角度提出如何扶持处在"瓶颈期"的企业,实现企业在数量、规模和质量上的同步提升。研究结果表明,Logistic回归模型可以更有效地预测县域中小企业生存状况;不同的注册资本、产业类别、企业类别和企业类型对县域中小企业生存时间存在显著性影响,注册资本越高,生命周期越长;第三产业的企业生存时间更长;国有企业比其他四种企业生存时间更长;农专生存概率最高,内资企业生存概率最低。
We compare and analyze the relationship between the survival time of small and medium-sized enterprises in China and their registered capital, industry category, enterprise category and enterprise type, with small and medium-sized enterprises as sample data, by adopting the Logistic regression model, the decision tree model and the k-nearest neighbor classification model. From the perspective of the government and enterprises, we show how to support enterprises in the"bottleneck period" and achieve the simultaneous improvement in the quantity, scale and quality of enterprises. The results show that the Logistic regression model can predict the survival of small and medium-sized enterprises in the county more effectively. Different registered capital, industry categories, enterprise categories and enterprise types have significant influence in the survival time of small and medium-sized enterprises. It is found that the higher the registered capital results in the longer the life cycle, the longer the survival time of the enterprises in the tertiary industry, the longer the survival time of the state-owned enterprises than the other four enterprises, the highest survival probability of the agricultural specialties, and the lowest survival probability of the domestic enterprises.
引文
[1] ABDESSELAM R,BONNET J,LE PAPE N.An explanation of the life Span of new french firms[J].Small Business Economics,2004,23(3):237-254.
[2]陈晓红,曹裕,马跃如.基于外部环境视角下的我国中小企业生命周期——以深圳等五城市为样本的实证研究[J].系统工程理论与实践,2009,29(1):64-72.
[3]张静,胡倩,谭桑,等.进入、退出与企业生存——来自中国制造业企业的证据[J].宏观经济研究,2013(11):103-143.
[4]“中国企业寿命测算方法及实证研究课题组”,何平.企业寿命测度的理论和实践[J].统计研究,2008(4):20-32.
[5]郑建贞,刘秀玲.中国企业OFDI持续时间的生存分析[J].国际商务,2014(5):66-73.
[6]吴喜之.应用回归及分类[M].北京:中国人民大学出版社,2016:78-79.
[7]李航.统计学习方法[M].北京:清华大学出版社,2012:100-103.
[8]韩家炜.数据挖掘:概念与技术[M].范明译.北京:机械工业出版社,2012:275-277.
[9]魏瑾瑞,吕晓云.Logistic模型对非平衡数据的敏感性:测度、修正与比较[J].统计研究,2016,33(2):79-85.
[10]孔志周,蔡自兴.多分类器融合框架下的模糊积分方法[J].统计与决策,2007(10):137-139.
[11]李太顺,刘沛.ROC曲线绘制和曲线下面积比较的SAS宏包[J].中国卫生统计,2018,35(2):302-304.
[12]吕光明.中国货币政策产业非均衡效应实证研究[J].统计研究,2013,30(4):30-36.