基于SVM的国防科研院所自主创新能力预测模型
The Independent Innovation Capability Predication Model of National Defense Research Institutes Based on SVM
-
摘要: 在基于自主创新的过程与结果互为转化的基础上,将指标分为动态参数层和静态参数层,每 个层又细分为要素组和指标集合,从而构建了多层次的指标评估体系。在该体系中引入支持 向量机(SVM,Support Vector Machine)方法建立系统的评价预测模型,将获取的科研院 所指标集分为训练样本集和检验样本集,经过训练样本集训练,检验样本集进行检验后,即 可对输入的指标集进行其自主创新能力预测。Abstract: Based on commutative transform between the process and result of independent innovation, the indexes of this paper are divided into the dynamic parameter layer and the static parameter layer. Each layer is divided into element groups and index gathers. Then the multi-hierarchy indexes evaluation system is established. It introduces the method of SVM, Support Vector Machine to establish the evaluation and predication model in the system. It works by training and verifying the sample gather and the capability of testing sample is predicated.