论文部分内容阅读
供应商选择是一个多影响因素、多层次的决策问题。为了选择很好的供应商,需要对影响因素进行特征选择和对供应商进行有效的分类。目前有效的同时解决这两个问题的方法比较欠缺。将P-SVM方法应用于供应商选择模型中,其中微分进化算法和遗传算法应用于参数的优选。试验结果显示在P-SVM参数优选上,微分进化算法比遗传算法更具有稳定性,数据仿真显示P-SVM方法比标准支持向量机(SVM)方法有更高的分类精度和推广能力。
Supplier selection is a multi-factor, multi-level decision-making problem. In order to choose a good supplier, the selection of the influencing factors and the effective classification of the suppliers are needed. The current effective solution to both these problems is lacking. The P-SVM method is applied to the supplier selection model, in which the differential evolution algorithm and the genetic algorithm are applied to the optimization of the parameters. The experimental results show that the proposed algorithm is more stable than the genetic algorithm in the P-SVM parameter optimization. The data simulation shows that the P-SVM method has higher classification accuracy and promotion ability than the standard support vector machine (SVM) method.