Data mining of corporate financial fraud based on neural network model

Accounting Department, Business School, Changchun Guanghua University ,S.L Li

doi:10.18287/2412-6179-co-656

Accounting Department, Business School, Changchun Guanghua University , S.L Li

Open Access

https://doi.org/10.18287/2412-6179-co-656

Copy DOI

Journal: Computer Optics	Publication Date: Aug 1, 2020
Citations: 6	License type: cc-by

Abstract

Under the active market economy, more and more listed companies emerge. Because of the various interest relationships faced by listed companies, some enterprises which are not well managed or want to enhance company’s value will choose to forge financial reports by improper means. In order to find out the false financial reports as accurately as possible, this paper briefly introduced the relevant indicators for judging the fraudulence of financial reports of listed companies and the recognition model of financial reports based on back propagation (BP) neural network. Then the selection of the input relevant indexes was improved. The improved BP neural network was simulated and analyzed in MATLAB software and compared with the traditional BP neural network and support vector machine (SVM). The results showed that the importance of total assets net profit, earnings per share, cash reinvestment rate, operating gross profit and pre-tax ratio of profit to debt was the top 5 among 20 judgment indexes. In the identification of testing samples of financial report, the accuracy, precision, recall rate and F value all showed that the performance of the improved BP neural network was better than that of the traditional BP network and SVM.

Highlights

Since the reform and opening up, China’s market economy has gradually become active, the growth rate is increasing day by day, and more and more companies are listed on the stock market [1]
Kanapickienė et al [6] identified whether financial reports were fraudulent or not based on financial ratio analysis, made financial ratio analysis on 40 sets of fraudulent financial reports and 125 sets of real financial reports using logistic regression model, and found that the method could effectively identify the fraudulent financial reports
This paper briefly introduces the relevant indicators used for determining the fraudulent financial reports of listed companies and the recognition model of financial reports based on Back Propagation (BP) neural network and gives a simulation analysis on the regular financial reports and irregular financial reports in CSMAR database by using MATLAB software

Summary

Introduction

Since the reform and opening up, China’s market economy has gradually become active, the growth rate is increasing day by day, and more and more companies are listed on the stock market [1]. In the actual operation process, some companies will make false statements on financial reports for some economic. If the error is within the specified range, the result will be output directly; If not, the weight and bias terms of the calculation formula in the hidden layer and the output layer will be reversely adjusted. The process from the forward calculation to the reverse adjustment of weight and bias terms according to error is considered one time of iteration. The traditional BP neural network described above can effectively fit the change rule to a certain extent in the training of financial report authenticity identification, there are 20 identification variables to be input in the model training as described above, which are relatively large in quantity. A report is determined as real if the value is smaller than 0.5; otherwise it is determined as fake

Experimental environment

Experimental data

Experiment setup

Criteria for judging the recognition effect of model

Experimental results

Conclusion