基于高纬稀疏数据分析的制造型企业研发投入影响因素分析

What Influence the R&D Intensity of Manufacturing Enterprises: the Research by Using Big Data

摘要:

国内外学者对影响企业研发投入的影响因素进行了大量研究和实证工作，现有研究大多是对影响企业研发投入的某一方面或几个方面及其交互效应进行定性研究，研究结论多以定性为主。有少量学者用定量分析的方法表达某个或者某几个变量与研发投入之间的数学关系，但由于科研投入的影响变量太多，这种定量分析方法的局限性大。为探索一种对大量因素同时进行定量分析的方法，尝试采用对高纬稀疏数据进行分析，对沪深两市制造型企业进行分析，以研发投入强度达到3.5%以上为分析目标，并得到达到这一分析目标的定量优化规则：当收入熵小于0.5794、销售毛利率适中（17.5%-29.54%）、董事职能背景差异小于0.3995、前10大股东持股比例适中（0.1689-0.6657），企业研发投入强度达到分析目标的比例会大幅提升。这一分析方法既可以用于国有高新技术企业经营层、董事会人员配置作为参考，又可作为对企业研发投入行为的预判。

Abstract:

Large amount of research and empirical work have been done on the influencing factors of R&D investment by domestic and international scholars. However, most of the existing works focus on qualitative research of one or several aspects and the interactive effects that affect the R&D investment of enterprises, which mainly came up with qualitative conclusion. A small number of scholars applied quantitative analysis to express the mathematical relationship between one or several variables and R&D investment, but in terms of the large number of influencing factors, this quantitative analysis method has its own limitation. In order to explore a method of quantitative analysis of a large number of factors at the same time, this paper tries to analyze the manufacturing enterprises in stock markets of Shanghai and Shenzhen by using the analysis method of big data. The analysis target is set as more than 3.5% R&D input intensity to get quantitative optimization rules.