粗集-遗传支持向量机模型在供应链绩效评价中的应用

Application of Rough Set GA-SVM Model in the Supply Chain Performance Evaluation

将粗集-遗传支持向量机模型运用到供应链绩效评价中,首先利用粗集理论剔除影响供应链绩效评价的冗余因素,获得核心影响因素,再采用支持向量机对于提取得到的核心影响因素预测供应链绩效所处的级别。在支持向量机分类过程中,利用遗传算法对支持向量机算法的参数进行寻优,获得最佳参数模型,而后预测得到供应链绩效评价级别。最后,实例运用此模型进行了预测,并与只运用粗集-支持向量机进行预测的结果进行对比。结果表明,利用粗集-遗传支持向量机方法对供应链绩效评价级别的预测准确率更高,预测结果更符合实际,是一种科学可行的方法。

Rough set-genetic support vector machine（SVM） model was used for the performance evaluation of supply chain.Firstly, this paper used rough set theory to eliminate the redundant factors influencing the performance evaluation of supply chain, and got the core factors. Support vector machine（SVM） was then used to extract the core level of influing factors to predict the performance of the supply chain. In the process of support vector machine（SVM） classification, the genetic algorithm was used for the parameters optimization of support vector machine（SVM） algorithm. Optimal parameters of the model were obtained and then used to predict the level of supply chain performance evaluation. Finally, instances were forecasted by this model, and compared with the predic results by the only use of rough sets and support vector machine（SVM）. The results showed that the use of rough set-genetic support vector machine（SVM） method can predict higher accuracy level of the supply chain performance evaluation, and the predicted results are more realistic, thus being a scientific and feasible method.