Evaluation of aviation equipment supplier based on BP-Adaboost algorithm and TOPSIS
顾玉磊 1马晖 2王愚勤 1胡卉 2刘富鑫1
扫码查看
点击上方二维码区域，可以放大扫码查看
作者信息
1. 长安大学汽车学院,陕西 西安 710064
2. 长安大学运输工程学院,陕西 西安 710064
折叠
摘要
为提升航空装备供应商评价标准的适用性及准确性,提出并设计出一种基于逆向传播(back propagation,BP)神经网络-弱分类器(Adaboost)与优劣解距离法(technique for order preference by similarity to an ideal solution,TOPSIS)的航空装备供应商评价方法.运用因子分析理论对供应商评价指标进行筛选,构建适合航空装备行业特点的供应商评价指标体系.在Adaboost算法元框架下,将BP神经网络作为基分类器,设计基于BP-Adaboost强分类器供应商分类模型.针对BP-Ada-boost算法无法精确计算供应商综合得分的不足,设计基于TOPSIS法的供应商评价模型.案例分析结果表明,基于 BP-Adaboost与TOPSIS法的航空装备供应商评价模型具有更高的评价准确度,对企业完善供应商管理体系提供理论和实践指导.
Abstract
In order to improve the applicability and accuracy of the evaluation criteria for aviation equipment suppliers,a supplier evaluation method was proposed and designed based on back propagation(BP)neural network-weak classifier(Adaboost)and technique for order preference by similarity to an ideal solution(TOPSIS).In the process of constructing the evaluation model,the supplier evaluation indexes were selected by factor analysis theory,and a supplier evaluation index system was established which was suitable for the characteristics of the aviation equipment industry.Being taken BP neural network as base classifier,a supplier classification model based on strong classifier of BP-Adaboost was designed under the Adaboost algorithm framework.In view of the supplier's comprehensive score could not be accurately calculated by BP-Adaboost algorithm,a supplier evaluation model was designed based on TOPSIS method.The case results showed that supplier evaluation model based on BP-Adaboost and TOPSIS had higher evaluation accuracy,which provided theoretical and practical guidance for enterprises to perfect supplier management system.
维普
万方数据