基于贝叶斯网络的性别失衡风险传导研究

Study on Gender Imbalanced Risks Conduction Based on Bayesian Networks

性别结构失衡已成为当前中国重要的人口问题之一。作为一种社会风险,其影响正通过特定途径扩展到社会其他领域,引发包括社会安全在内的一系列社会问题。本文基于风险传导理论和性别失衡问题的相关研究成果,通过文献整理等方式构建出了一个性别失衡风险传导理论模型,并以该模型为框架,经过节点确定、结构学习和参数学习等步骤构建贝叶斯网络模型。进而以专家问卷调查数据为基础,从预测推理、诊断推理、影响强度分析和敏感性分析等四个方面对性别失衡风险传导问题进行了详细的探讨。

Gender imbalance has become one of the most important problems existing in the population domain.As a kind of the social risks,problems of gender imbalance have influenced other social domains by some paths and leaded to a series of problems including social security.Based on the risk conduction theory,theoretical model of the gender imbalance risk conduction is established in this paper.Then,based on definition of nodes,structure learning and the parameter learning,Bayesian network of the gender imbalance risk conduction is developed.In the end,by the questionnaire researches from decades of experts,issues on gender imbalanced risks conduction are argued from four aspects including predictive inference,diagnostic inference,and sensitivity analysis in this paper.Article Type: Research paper