Enhancing the detection of financial statement fraud through the use of missing value estimation, multivariate filter feature selection and cost-sensitive classification

Stephen O. Moepya

Back

Enhancing the detection of financial statement fraud through the use of missing value estimation, multivariate filter feature selection and cost-sensitive classification

Dissertation

Open access

Enhancing the detection of financial statement fraud through the use of missing value estimation, multivariate filter feature selection and cost-sensitive classification

Stephen O. Moepya

DPhil, University of Johannesburg

2017

Handle:

https://hdl.handle.net/10210/242812

Abstract

Data mining

Fraud - Statistical methods

Missing observations (Statistics)

Computational intelligence

D.Phil. (Electrical and Electronic Engineering) Please refer to full text to view abstract

Files and links (1)

pdf

Enhancing the detection of financial statement fraud through the use of missing value estimation, multivariate filter feature selection and cost-sensitive classificationDownload View

Open Access

Metrics

11 File views/ downloads

39 Record Views

Details

Title: Enhancing the detection of financial statement fraud through the use of missing value estimation, multivariate filter feature selection and cost-sensitive classification
Creators - without role: Stephen O. Moepya
Contributors - without role: B. Twala
F. Nelwamondo
Awarding Institution: University of Johannesburg; DPhil
Theses and Dissertations: DPhil, University of Johannesburg
Identifiers: 9913005107691
Copyright: University of Johannesburg
Academic Unit: Electrical and Electronic Engineering Studies
Resource Type: Dissertation