Font Size: a A A

Risk Early Warning Of Chinese Stock Market Based On Support Vector Machine After The Reform Era

Posted on:2015-12-27Degree:MasterType:Thesis
Country:ChinaCandidate:Q FuFull Text:PDF
GTID:2309330467965001Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
Along with the constant outburst of stock market crisis throughout the world,the correlation and destructiveness of the crisis are becoming more and moreobvious. Due to the practical relations between risks and crisis of stock market,the former risks tend to be the prelusion, or beginning of the latter; while the latteris the extreme consequence of dynamic accumulated the former. Therefore, theresearch on early warnings of risks of stock has caused widespread governmentaland public concern from all around the world, the results of which will directlyaffect the proper understanding and judgment of financial market status, thusbringing methods to prevent the risks in the stock market from taking place,withstanding the impacts of international financial crisis and regulating thefinancial risks to be within the controllable range. It has great practicalsignificance to the healthy, stable development of the whole economy in ourcountry.This article is based on the theories of stock market risk early warnings andSupport Vector Machines. In accordance with research findings in the area ofstock risk early warning index system by experts from home and abroad, itchooses the index of the risk early warnings appropriate for the practical situationin China after comparatively sufficient acknowledgments of the present situationand research findings in this field. Constantly constrained by the experience ofexperts and simple mathematical models, traditional early warning methods arenot sufficient in coping with highly nonlinear models, thus not fulfilling theobjective requirements macroscopic early-warning of finance. This article usesSupport Vector Machines and its deformation algorithmic model to testify thefeasibility and rationality of the selected early-warning index system andsuccessfully predicts the risks of stock market in upcoming years in China and proves the feasibility of this approach.
Keywords/Search Tags:early warning of stock market risks, Statistical Learning Theory, Support Vector Machine (SVM), Support Vector Classification (SVC)
PDF Full Text Request
Related items