ResearchSpace

A support vector machine approach to detect financial statement fraud in South Africa: A first look

Show simple item record

dc.contributor.author Moepya, SO
dc.contributor.author Nelwamondo, Fulufhelo V
dc.contributor.author Van der Walt, C
dc.date.accessioned 2014-07-30T09:14:58Z
dc.date.available 2014-07-30T09:14:58Z
dc.date.issued 2014-04
dc.identifier.citation Moepya, S.O, Nelwamondo, F.V and Van der Walt, C. 2014. A support vector machine approach to detect financial statement fraud in South Africa: A first look. In: 6th Asian Conference on Intelligent Information and Database Systems, Bangkok Thailand, 7-9 April 2014 en_US
dc.identifier.uri http://link.springer.com/chapter/10.1007%2F978-3-319-05458-2_5
dc.identifier.uri http://hdl.handle.net/10204/7532
dc.description 6th Asian Conference on Intelligent Information and Database Systems, Bangkok Thailand, 7-9 April 2014 en_US
dc.description.abstract Auditors face the difficult task of detecting companies that issue manipulated financial statements. In recent years, machine learning methods have provided a feasible solution to this task. This study develops support vector machine (SVM) models using published South African financial data. The input vectors are comprised of ratios derived from financial statements. The three SVM models are compared to the k-Nearest Neighbor (kNN) method and Logistic regression (LR). We compare the ability of two feature selection methods that provide an increase classification accuracy. en_US
dc.language.iso en en_US
dc.publisher Springer link en_US
dc.relation.ispartofseries Workflow;13030
dc.subject Financial statement fraud detection en_US
dc.subject Machine learning en_US
dc.subject Support vector machines en_US
dc.subject Logistic regression en_US
dc.subject LR en_US
dc.title A support vector machine approach to detect financial statement fraud in South Africa: A first look en_US
dc.type Conference Presentation en_US
dc.identifier.apacitation Moepya, S., Nelwamondo, F. V., & Van der Walt, C. (2014). A support vector machine approach to detect financial statement fraud in South Africa: A first look. Springer link. http://hdl.handle.net/10204/7532 en_ZA
dc.identifier.chicagocitation Moepya, SO, Fulufhelo V Nelwamondo, and C Van der Walt. "A support vector machine approach to detect financial statement fraud in South Africa: A first look." (2014): http://hdl.handle.net/10204/7532 en_ZA
dc.identifier.vancouvercitation Moepya S, Nelwamondo FV, Van der Walt C, A support vector machine approach to detect financial statement fraud in South Africa: A first look; Springer link; 2014. http://hdl.handle.net/10204/7532 . en_ZA
dc.identifier.ris TY - Conference Presentation AU - Moepya, SO AU - Nelwamondo, Fulufhelo V AU - Van der Walt, C AB - Auditors face the difficult task of detecting companies that issue manipulated financial statements. In recent years, machine learning methods have provided a feasible solution to this task. This study develops support vector machine (SVM) models using published South African financial data. The input vectors are comprised of ratios derived from financial statements. The three SVM models are compared to the k-Nearest Neighbor (kNN) method and Logistic regression (LR). We compare the ability of two feature selection methods that provide an increase classification accuracy. DA - 2014-04 DB - ResearchSpace DP - CSIR KW - Financial statement fraud detection KW - Machine learning KW - Support vector machines KW - Logistic regression KW - LR LK - https://researchspace.csir.co.za PY - 2014 T1 - A support vector machine approach to detect financial statement fraud in South Africa: A first look TI - A support vector machine approach to detect financial statement fraud in South Africa: A first look UR - http://hdl.handle.net/10204/7532 ER - en_ZA


Files in this item

Thumbnail
Name: Moepya_2014_ABSTRACT ...
Size: 59.76Kb
Format: PDF
View/Open

This item appears in the following Collection(s)

Show simple item record