Font Size: a A A

Research On Construction And Application Of Financial Crisis Warning Model Based On Bayesian Network

Posted on:2018-06-28Degree:MasterType:Thesis
Country:ChinaCandidate:X L LiangFull Text:PDF
GTID:2359330542980331Subject:Asset Assessment
Abstract/Summary:PDF Full Text Request
Since Fitzpatrick first proposed the concept of financial crisis in 1932,corporate financial crisis warning has been the focus of academic research.From the definition of financial crisis,to the construction of influencing factors index system,and then to use data mining and other cross-domain knowledge to build financial early warning model,domestic and foreign scholars are seeking closer to the practical application of the financial crisis early warning model.With the deepening of China's comprehensive reform of the system and the global economy continued to weaken the macro factors such as China's enterprises are facing more rapid changes in the risk of a more diverse environment,which for enterprises to identify enterprises in a timely manner to identify potential financial risks Make the corresponding measures put forward higher requirements.This paper,through combing domestic and foreign research,finds out the shortcomings of the current index system construction,breaks through the traditional index system with the financial ratio as the theme,and constructs the perfect financial early warning index system according to the macro factors,the governance structure factor and the stakeholder theory.In this paper,based on the knowledge of data mining,this paper proposes a Bayesian network method based on Bayesian theory.Bayesian network approach,as a machine learning algorithm combining probability theory and graph theory knowledge,has a very good performance in dealing with uncertainty,and its advantages in dealing with qualitative discretization indicators make it better suited to join Macro factors,governance structure and other qualitative indicators of the index system.At the same time,in order to test the prediction ability of the early warning model,this paper examines the accuracy of the model by using the logit model as a comparative model,and then applies the model to the GN enterprise as a case to further test the forecasting ability of the early warning model.In summary,this paper constructs the enterprise financial crisis early warning model with certain forecasting ability,at the same time for the stakeholder theory and Bayesian network method application and enterprise financial early warning to make certain development,has the theory and the practical significance.
Keywords/Search Tags:Financial Crisis, Financial Early Warning, Stakeholder Theory, Bayesian Network
PDF Full Text Request
Related items