基于主成分分析-支持向量机的我国农业科技创新能力评价

Evaluation of Innovation Ability of Agricultural Science and Technology in China Based on Principal Component Analysis and Support Vector Machine

针对目前我国在农业科技创新能力评价中时间维度研究缺失、指标选取过多、指标相关性不足的现状,该研究采用主成分分析法,将影响农业科技创新能力的众多因素进行降维,从而找出重要影响因素作为支持向量机模型的输入层,建立精度更高、数据需求量更少、计算时间更短的农业科技创能力评价模型。运用该模型得到我国2005—2016年农业科技创新能力水平,得出农业科技创新能力由投入产出以及研发能力共同决定。最后,根据该研究结果提出提升我国农业科技创新能力的相关建议。

In view of the current situation of lack of time dimension research,excessive selection of indicators and insufficient correlation of indicators in the evaluation of agricultural science and technology innovation ability in China,we used principal component analysis (PCA) to reduce the dimensionality of many factors affecting the innovation ability of agricultural science and technology,to find out the important factors as the input layer of SVM model,establish the evaluation model of agricultural science and technology innovation ability with higher accuracy,less data demand and shorter calculation time.This model was used to obtain the level of agricultural science and technology innovation ability in China from 2005 to 2016.According to the results,it is found that the innovation ability of agricultural science and technology is determined by the input and output as well as the research and development ability.Finally,according to the results of this study,some suggestions are put forward to improve the innovation ability of agricultural science and technology in China.