基于支持向量机和神经网络的供应商选择方法比较

The comparison on neural network and support vector machine in supplier selection

供应商选择是供应链管理的重要内容,近年来吸引大量学者进行研究,其中大量文献显示神经网络方法比传统统计方法有更大的优越性。然而神经网络具有固有的缺陷,如最优解的局部性、泛化能力低、训练样本大和无法控制收敛等。引用新的机器学习技术---支持向量机(support vector machines,SVM),用于选择理想供应商,并与BP神经网络算法相比较。实证表明,支持向量机算法比神经网络算法计算精确。

Supplier selection in supply chain management has attracted lots of research interests in recent years. Recent literatures have shown that neural network achieved better performance than traditional statistical methods. However, neural networks have inherent defects, such as locally optimal solution, worse generalization, finite samples and uncontrolled convergence. In this paper, a relatively new machine learning technique, support vector machines （SVM）, is introduced to provide a model with better explanatory power to select ideal supplier partners. We used backpropagation neural network （BNN） as a benchmark and compare prediction accuracy for both BNN and SVM methods for the supplier selection, the actual examples illustrate that SVM methods are superior to NN methods.