基于SVM的建筑企业信用评价研究

To Study on Credit Evaluating of Construction Enterprises Based on SVM

建筑业出现的严重信用危机成为亟待解决的问题。在建立了初始建筑企业评价指标体系,对各评价指标进行相关性分析的基础上,建立最终的评价指标体系。采用基于统计学理论和结构风险最小化准则的支持向量机(SVM)方法进行信用评价,选用径向基核函数,使用10-折交叉检验,grid搜索确定最优参数。在此基础上建立建筑企业信用评价模型,并对其预测准确性进行分析。

The serious credit crisis in construction industry, which is a pillar industry in national economy, has become a key problem, it faces serious credit crisis. Based on systematic analysis of the credit index system, a preliminary credit indexes was established, On the basis of correlation analysis the final credit index system was proposed. SVM, which is based on the Structural Risk Minimization principle, was choosed to evaluate the construction enterprise credit. To build the credit evaluation model, Radial Basis kernel Function was selected as Support Vector kernel function and Grid-search for the best parameter under the 10-fold cross-validation procedure was adopted. The credit prediction accuracy was analyzed in the end.