中国专利活动投入要素的产出弹性分析——基于状态空间模型的变参数估计  

Analysis the output elasticity of input factors of patent activities in China——the time varying parameter estimating based on state space model

作　　者：杜鹃[1] 陶磊[1] 胡敏杰[1] 

机构地区：[1]西南交通大学经济管理学院,四川成都610031

出　　处：《科学学研究》2008年第5期948-954,共7页Studies in Science of Science

基　　金：教育部高等院校博士点专项科研基金(20060613019)

摘　　要：专利产出水平体现了一国的科技创新能力,对国家经济发展的贡献份额越来越大,在追求最优的专利投入产出效益过程中,对专利投入产出的定量分析必不可少。与以往文献注重静态分析不同,本文首次分析了专利活动的投入与产出之间的动态关系。本文在知识生产函数理论的基础上构建了专利生产函数模型,在多变量分析框架下,将其设定成基于状态空间形式的变参数模型,根据Kalman滤波法,运用一组时间序列数据对参数进行连续地估计,并对其波动进行了分析。Patent output of a country is the embodiment of its technology innovation capability, contribution of which to economic growth is more and more big. In the process of seeking the best benefit for patent input and output, a quantitative analysis on input - output of Chinese patent is absolutely necessary. Contrast to the former literature that pay attention to the static or balanced analysis, this text analyzed dynamic relationship and unbalanced relationship between the input and output of patent activity for the first time. A patent produce model is built abased on theory of knowledge produce in this text, which is constructed the time varying parameter model based on state space model Under the analysis frame of multi - variables. Then continuous calculation is realized for the parameters of state space model by KALMAN filter arithmetic and analyse their size and fluctuation.

关 键 词：专利活动 投入产出 状态空间模型 变参数分析 

分 类 号：F062.4[经济管理—政治经济学]