Document Type : Research Paper
Authors
1 Phd Student, Department of Accounting, Borujerd Branch, Islamic Azad University, Borujerd, Iran
2 Associate Professor, Department of Accounting, Borujerd Branch, Islamic Azad University, Borujerd, Iran
3 Assistant professor of computer Department, faculty of Engineering Shahid Chamran university of Ahvaz. Ahvaz, Iran
4 Assistant Professor, Department of Statistics, Faculty of Science, Fasa University, Fasa, Iran
Abstract
The purpose of this research is modeling the detection of firms financial fraud under the implementation of artificial neural network's evaluation algorithms. In this study, efforts have been made by using Quadratic Programming "QP" processes in artificial neural network algorithms to determine the basic algorithm in the first place and choose the technical parameters of the artificial neural network in the second place, based on the time data from 2013 to 2022, through several stages. Then, by developing a diagnostic model based on two test and control scales, innovative algorithms that have the highest accuracy coefficients in predicting the accuracy of financial fraud should be investigated at the level of capital market companies. Therefore, based on the systematic sampling process, 95 stock exchange companies were selected, so that based on 950 observations (company-year), the distance between companies with financial health and companies with the possibility of financial fraud was determined through decimalization and the companies placed in the deciles with financial fraud should be examined through the parameters of the artificial neural network's usefulness. The results of the study showed that the unsupervised learning algorithm, which includes a set of evaluation parameters based on meta-heuristic algorithm, has higher accuracy of predictions based on the fulfilled data. Also, the results of predicting the financial frauds of decimated companies based on two selected algorithms, genetic and bee colony, show that the bee colony algorithm has a higher accuracy factor in predicting the probability of fraud of the investigated companies.
Keywords
Main Subjects
)