Research On Hybrid Model Of Recognizing Financial Reporting Fraud In Manufacturing Enterprises

Financial reporting fraud is a major performance in distortion of accounting information, which is manifested to forgery business accounting fraud or to report accounting information in a wrong or omission way in conscious. Financial reporting fraud has already started to occur since three hundred years ago. With the continuous development of the economy in human society, financial reporting fraud is constantly emerging worldwide, which is seriously hampering the sustainable development of healthy and stable economy. As a result, research on the model of recognizing financial reporting fraud has a very important practical significance. Because the same industry often has their own economic characteristics, in order to achieve the target of prediction as well as testing accurately and to get the samples conveniently, this article studies the hybrid model of recognizing financial reporting fraud with manufacturing enterprises as the samples.In current academic field, there have been some scholars focusing on financial reporting fraud, and also giving some recognition models, but each model has its own advantages and disadvantages. Through my studies, it is found that using Logistic regression model and BP neural network technology to build a new hybrid model can not only overcome each model's disadvantages but also improve the accuracy of recognition model. Therefore, we use Logistic regression models and BP neural network method of artificial neural network technology to establish a new hybrid models. In addition, in order to improve the testing quality of this model of recognizing the financial reporting fraud, we use both the generation of test samples and the prediction of test samples to test such model's recognizing capability.