基于变权Markov链的中小物流企业信用风险集对预测  

Set Pair Credit Risk Predication Model for Small and Medium-sized Logistics Enterprises Based on Variable Markov Chain

作　　者：刘若阳[1] 申威 唐长虹[1] LIU Ruo-yang;SHEN Wei;TANG Chang-hong(School of Logistics,Beijing wuziUniversity,Beijing 101149,China)

机构地区：[1]北京物资学院物流学院,北京101149

出　　处：《系统工程》2021年第2期138-149,共12页Systems Engineering

基　　金：2018北京市自然科学基金面上项目(4192020);2018年北京社科基金项目研究基地一般项目(18JDGLB025)。

摘　　要：中小物流企业信用表现的优劣直接关系物流服务供应链生态系统的平衡性和稳定性。针对现有研究仅侧重历史信用评估,无法有效判断企业未来风险演变的情况,本文构建集对-变权Markov模型预测中小物流企业未来信用风险趋势。首先,构建完善的中小物流企业信用特征指标体系,设计CRITIC-时间熵方法综合面板信用数据纵横向信息计算指标变权;其次,综合考虑风险状态转移和趋势转移两种情景分别构造集对关系和M a rko v转移矩阵,结合集对势思想从正、均、反三种态势预测企业信用发展;最后,通过实例分析验证模型对于中小物流企业信用风险状态和趋势判断的有效性。该模型从未来风险发展视角分析中小物流企业信用状态,为有效预判物流企业信用趋势提供决策支持。Considering the existing research only focuses on the historical credit evaluation,which is unable to predict the future risk evolution of small and medium logistics enterpriseseffectively,and the credit ecological imbalance in the logistics industry is becoming increasingly severe,this paper constructs a Pair-Variable Markov model.Firstly,a CRITIC-Time entropy method combined with vertical and horizontal information is constructed to calculate the dynamic weights for the credit index.Secondly,risk state transition and trend transition are proposed to construct set-pair relationships and transition matrices, and the set-pair potential index is used to predict the overall credit from the positive,average, and negative situations,which has improved the original theory.Finally, the case shows the effectiveness of this model,which can fully mine the credit data of the panel information,and has a better early warning of the enterprises’ credit risk trend.

关 键 词：中小物流企业 信用风险 MARKOV链 集对分析 预测 

分 类 号：F259[经济管理—国民经济]