摘要
在国内外的研究基础上,结合安徽特有的基本情况,根据安徽省"1+6"政策体系,建立了一套高新技术统计指标体系。以最大依赖性、最大相关性和最小冗余为准则建立模型,选择过滤式特征选择方法的代表算法之一mRMR来选择特征子集,在众多指标中抽取关键指标,并利用数据挖掘中聚类分析方法挖掘指标间潜在的关联性,提出高新技术产业增加值和高新技术企业培育情况是影响一个地区高新技术产业运行情况的重要指标。
Firstly,on the basis of research at home and abroad,and combining the basic situation of Anhui characteristics,this article establishes a set of index system of new and high technology industries.Secondly,to maximize the dependency,maximum correlation and minimum redundancy for the guidelines,this article establishes a model,chooses mRMR to select feature subset which is one of the representative algorithms of the filter,and extract the key indexes in many indexes.Thirdly,data mining the potential correlation between excavated index using the method of clustering analysis.At last,put forward that it is the added value of new and high technology industries and the enterprises which affect the high and new technology industry.
作者
王俊
WANG Jun(Scientific and Technological Information Institute of Anhui Province, Hefei 230011)
出处
《中国科技资源导刊》
2017年第2期88-92,共5页
China Science & Technology Resources Review
基金
安徽省科技攻关计划项目"高新技术统计关键指标挖掘研究"(1301023012)
国家创新发展司委托项目子课题"安徽省企业创新情况调查分析与研究"(ZLY2015123)
关键词
高新技术
数据挖掘
关键指标
相关度
安徽省
new and high technology
data mining
key indexes
relativity
Anhui province