基于投影寻踪-随机森林的制造企业供应链稳定性评价

Stability Evaluation of Manufacturing Enterprise Supply Chain Based on Projection Pursuit and Random Forest Method

基于扎根理论提取制造企业供应链稳定性影响因素,鉴于因素的随机性和模糊性特征,引入云模型,将定性概念转化为定量数据,从众多影响因素中筛选出重要因素,构建制造企业供应链稳定性评价指标体系。采用加速遗传算法对建立的投影寻踪模型进行求解,确定指标权重与稳定性评价值。利用随机森林方法根据评价值划分供应链稳定性类型,检验评价方法科学性,并确定当前制造企业供应链稳定性的整体状态,据此提出优化制造企业供应链稳定性建议。

Based on the grounded theory,the influencing factors of supply chain stability of manufacturing enterprises are extracted.According to the factors’characteristics of randomness and fuzziness,the cloud model method is introduced to realize the transformation from qualitative concept to quantitative data.The important influencing factors are selected from several influencing factors,and the evaluation index system of supply chain stability of manufacturing enterprises is constructed.Using projection pursuit model based on accelerated genetic algorithm,the index weight and stability evaluation score of manufacturing enterprises are confirmed.Based on the evaluation score,the random forest method is used to classify the types of supply chain stability and test the scientific nature of the evaluation method,and determine the overall condition of the current manufacturing enterprise supply chain stability,and accordingly the suggestions for optimizing the supply chain stability of manufacturing enterprises are brought forward.