企业运行指标因果分析的动态贝叶斯网络方法

针对现有的企业运行指标分析方法只强调动态或静态信息,不易实现二者结合的情况,建立了用于企业运行指标因果分析的动态贝叶斯网络模型,这种模型可将时间片间的指标动态时序因果关系与时间片内指标静态因果联系融为一体,并通过量化推理进行动态与静态因果分析。通过与领域专家交流,所建立的企业运行指标动态贝叶斯网络良好地反映了数据中所蕴涵的因果关系。...

Full description

Saved in:
Bibliographic Details
Published in计算机应用研究 Vol. 33; no. 5; pp. 1433 - 1436
Main Author 高瑞 王双成 杜瑞杰
Format Journal Article
LanguageChinese
Published 上海立信会计学院数学与信息学院，上海201620 2016
上海财经大学 统计与管理学院，上海200433%上海立信会计学院数学与信息学院，上海201620
上海立信会计学院立信会计研究院，上海201620%上海立信会计学院数学与信息学院,上海,201620
Subjects
企业运行指标
动态贝叶斯网络
因果分析
马尔可夫毯
causal analysis
dynamic Bayesian networks
企业运行指标
Markov blanket
因果分析
马尔可夫毯
enterprise operation indexes
动态贝叶斯网络
Online AccessGet full text

Cover

More Information
Summary:针对现有的企业运行指标分析方法只强调动态或静态信息,不易实现二者结合的情况,建立了用于企业运行指标因果分析的动态贝叶斯网络模型,这种模型可将时间片间的指标动态时序因果关系与时间片内指标静态因果联系融为一体,并通过量化推理进行动态与静态因果分析。通过与领域专家交流,所建立的企业运行指标动态贝叶斯网络良好地反映了数据中所蕴涵的因果关系。
Bibliography:51-1196/TP
enterprise operation indexes; causal analysis; dynamic Bayesian networks; Markov blanket
Gao Rui;Wang Shuangcheng;Du Ruijie（ 1. a. School of Mathematics ＆ Information, b. Lixin Accounting Research Institute, Shanghai Lixin University of Commerce, Shanghai 201620, China; 2. School of Statistics ＆ Magagement, Shanghai University of Finance ＆ Economics, Shanghai 200433, China）
In the light of those methods now available for analyzing enterprises operation indexes are emphasizing only dynamic or static information,and have not realized the combinations between those two kinds of different information. This paper set up a dynamic Bayesian network method for causal analysis among enterprises operation indexes. The model could combine dynamic time sequence and static causal relationships of panel data as a whole,could analyze both dynamic and static causal relationships through quantitative inference without the assumptions of liner causal relationships. Communicating with the experts in the relative field,t
ISSN:1001-3695
DOI:10.3969/j.issn.1001-3695.2016.05.035