基于SVM的高技术装备制造业供应风险预测模型

通过对高技术装备制造业与供应商合作中供应风险的相关因素进行识别，确定了供应风险预测的指标体系。运用支持向量机(support vector machine， SVM)等数据挖掘方法和Libsvm(a library for support vector machine)技术建立了供应风险的预测模型。案例研究表明，预测模型的均方误差、平均差、整群剩余系数和确定系数等评价指标皆显现出较好的模拟效果，且系统误差不显著。预测模型在高技术装备制造企业供应风险管理的实践中具有较好的适用性和准确度。

Supply risk prediction model of high-tech equipment manufacturing industry based on SVM

Relevant factors of supply risk between high-tech equipment manufacturing industry and suppliers are identified to set up the supply risk predicting index system. On this basis, the supply risk predicting model is established with support vector machine (SVM) approach and Libsvm (a library for support vector machine) technology. The case study shows that all the model’s evaluation coefficients such as mean square error (MSE), mean difference between measurement and simulation (MD), coefficient of residual mass (CRM) and coefficient of determination (CD) show good simulation result, and the system error is not significant. The supply risk predicting model has good suitability and accuracy in supply risk management.