行政事业单位资产管理绩效评价研究——基于支持向量机的方法

Research on the Evaluation of Assets Management in Administrative Institutions:Based on the SVM Method

在总结国内外现有资产管理绩效评价相关研究的基础上,分析、挖掘中国行政事业单位资产结构特点,提炼相应的评价指标体系,利用关联分析,对指标体系进行约简,并选取北京市438家行政事业单位作为样本集,对支持向量机模型进行训练和检验,最终构建有效评价模型。实验结果表明,基于支持向量机的评价模型具有较高的判别精度和判别效率,对于参数选择的灵敏程度较低,应用价值高。

In this paper, we focused on the characteristic of the asset management in China, and abstracted the index system of evaluating the performance of asset management. Then,we based on the SVM to construct the performance evaluation model of assets management in administrative institutions. The experimental results show that SVM -based model is more precise and more efficient than that of the existing method like BP neural network. At the same time, SVM - based has less sensitivity of parameter selection and has good application value.