Identification of Fraudulent Financial Reporting with RBF Probability Neutral Network Model
Liping Wang · Journal of Harbin University of Commerce · 2006
The paper uses the RBF(Radial Basis Function) Probability Neutral Network Model to establish a model of fraudulent financial reporting for the purpose of testing the methods used for fraudulent financial reporting.The model's sample is consisted of 144 cross section indexes of 54 listed company of which 16 companies is used as a testing sample.After training and studying,we got a distinguishing positive rate of 88.7% to sample model setting and 86.7% the testing sample.This is a large improvement to the predicting precisions,in comparison with the other liner methods.