基于PCA-RBF神经网络的PPP项目风险智能评价研究

发布时间：2018-01-31 23:32

本文关键词： PPP项目 PCA-RBF神经网络智能评价模型风险评估　出处：《科技管理研究》2017年14期 　论文类型：期刊论文

【摘要】：随着新一轮政府和社会资本合作(public—private—partnership,PPP)模式在基础设施领域的推广运用,对其风险评估也开始备受关注。鉴于PPP项目风险影响因素众多和传统评估方法过度依赖主观评价等问题,在构建PPP项目风险评价指标体系的基础上,通过将主成分分析(principal component analysis,PCA)技术降维并结合模糊综合评价结果,建立自适应的径向基神经网络(RBF)的智能风险评价模型,并以入库的浙江省发改委10个PPP项目为例进行实证检验,结果显示3个项目处于风险较小、6个项目处于风险不大、1个项目处于风险中等。PCARBF智能评价模型所训练预测的结果与模糊综合评价结论一致,且在一定程度上提高了计算效率、增强了评价的客观性,为PPP项目风险智能评估提供理论上的参考。
[Abstract]:With the new round of government and social capital cooperation public-private-partnership-PPPmodel in the field of infrastructure promotion and application. In view of the numerous risk factors of PPP project and the over-reliance on subjective evaluation of traditional evaluation methods, based on the construction of PPP project risk evaluation index system. The principal component analysis (PCA) technique was used to reduce the dimension and combine with the result of fuzzy comprehensive evaluation. An intelligent risk assessment model of adaptive radial basis function neural network (RBFs) was established, and an empirical test was carried out on 10 PPP projects of Zhejiang Development and Reform Commission (NDRC). The results show that three projects are in a small risk, six projects are in a small risk, and one project is in a middle risk. PCARBF intelligent evaluation model training prediction results are consistent with the fuzzy comprehensive evaluation results. To a certain extent, the efficiency of calculation is improved, and the objectivity of evaluation is enhanced, which provides a theoretical reference for PPP project risk intelligent evaluation.
【作者单位】：嘉兴学院;东华大学;嘉兴职业技术学院;
【基金】：国家自然科学基金项目“面向服务智能协同的农业物联网动态自治与资源优化配置”(61473078) 浙江省自然科学基金项目“市场分割下FDI对区域资本流动的影响机理”(LY15G030034) 嘉兴学院南湖学院科研重点项目“浙江新型城镇化建设中PPP融资模式选择与风险评估研究”(N41472001-22)
【分类号】：F283;TP183
【正文快照】： 1研究背景政府和社会资本合作PPP(public—private—partnership)模式作为一种创新型的融资机制,在新型城镇化和“一带一路”背景下得到了广泛的应用,而其组织结构的特殊性伴随着其风险的复杂性,如何有效评价其风险便成为了当下各级政府部门、金融机构等所亟待解决的难题[1-2]

【相似文献】