摘要
基于第三次全国经济普查年鉴数据,运用机器学习算法和多元统计方法进行建模分析,构建了地区工业企业创新能力评价模型。首先利用K-means算法对31个省市自治区进行聚类,得到三个不同水平的分类;采用随机森林算法筛选影响类别划分的重要指标,其次运用因子分析方法提取影响地区工业企业创新能力的三个公因子,即资金投入、人力投入和创新成果;最后对各地区工业企业创新能力水平进行了综合分析,并对未来各地区工业企业创新发展提出建议。
On the basis of the third national economic census yearbook, this paper uses machine learning algorithm and multivariate statistical method to carry out modeling analysis, conducting an evaluation model of innovation ability for regional industrial enterprise. Firstly, we use K-means algorithm to classify 31 regions into three different levels;then random forest algorithm is used to screen the important indicators that affect classification greatly. In addition, we use factor analysis method to extract the three common factors of the industrial enterprises' innovation ability, namely, capital investment, human input and innovation results. Finally, the author conducts comprehensive analysis of industrial enterprises' innovation ability in various regions, and puts forward suggestions for the future development of industrial enterprises.
作者
陈艺曦
于博骏
CHEN Yi-xi;YU Bo-jun(School of Science, Beijing University of Posts and Telecommunication, Beijing 100876)
出处
《软件》
2019年第1期120-126,共7页
Software
关键词
地区差异
工业企业
创新能力
K-MEANS
随机森林
因子分析
Regional differences
Industrial enterprise
Innovation ability
K-means
Random forest
Factor analysis
