Zur Kurzanzeige

2008-01-07Diskussionspapier DOI: 10.18452/4096
The Default Risk of Firms Examined with Smooth Support Vector Machines
dc.contributor.authorHärdle, Wolfgang Karl
dc.contributor.authorLee, Yuh-Jye
dc.contributor.authorSchäfer, Dorothea
dc.contributor.authorYeh, Yi-Ren
dc.date.accessioned2017-06-15T23:36:18Z
dc.date.available2017-06-15T23:36:18Z
dc.date.created2008-01-10
dc.date.issued2008-01-07
dc.identifier.issn1860-5664
dc.identifier.urihttp://edoc.hu-berlin.de/18452/4748
dc.description.abstractIn the era of Basel II a powerful tool for bankruptcy prognosis is vital for banks. The tool must be precise but also easily adaptable to the bank's objections regarding the relation of false acceptances (Type I error) and false rejections (Type II error). We explore the suitability of Smooth Support Vector Machines (SSVM), and investigate how important factors such as selection of appropriate accounting ratios (predictors), length of training period and structure of the training sample influence the precision of prediction. Furthermore we show that oversampling can be employed to gear the tradeoff between error types. Finally, we illustrate graphically how different variants of SSVM can be used jointly to support the decision task of loan officers.eng
dc.language.isoeng
dc.publisherHumboldt-Universität zu Berlin, Wirtschaftswissenschaftliche Fakultät
dc.rights.urihttp://rightsstatements.org/vocab/InC/1.0/
dc.subjectInsolvency Prognosiseng
dc.subjectSVMseng
dc.subjectStatistical Learning Theoryeng
dc.subjectNon-parametric Classificationeng
dc.subject.ddc330 Wirtschaft
dc.titleThe Default Risk of Firms Examined with Smooth Support Vector Machines
dc.typeworkingPaper
dc.identifier.urnurn:nbn:de:kobv:11-10083063
dc.identifier.doihttp://dx.doi.org/10.18452/4096
local.edoc.pages32
local.edoc.type-nameDiskussionspapier
local.edoc.container-typeseries
local.edoc.container-type-nameSchriftenreihe
local.edoc.container-year2008
dc.identifier.zdb2195055-6
bua.series.nameSonderforschungsbereich 649: Ökonomisches Risiko
bua.series.issuenumber2008,5

Zur Kurzanzeige