Corporate Financial Fraud Identification and Crisis Forewarning Based on The Partial Least Squares Method
Conducting research into financial fraud and predicting financial crises is an important research issue. In this paper, we describe the development of a model based on partial least squares (PLS) combined with a support vector machine (SVM). Components were extracted through PLS and were used as the input of SVM. Then, data were distinguished using SVM. An analysis was carried out on the collected samples. The results show that after component extraction by PLS, the model achieved an average accuracy of 83.61% for fraud identification, and the model also achieved good performance in relation to financial crisis forewarning, with an accuracy of 95%, 90.51% and 88.45% for Year T-1, T-2 and T-3, respectively. The results verify the reliability of the method and that the method can be applied in practice.
Keywords: partial least squares, financial fraud identification, financial crisis warning, support vector machine
Y. Li, "Corporate Financial Fraud Identification and Crisis Forewarning Based on The Partial Least Squares Method",
Engineering Intelligent Systems, vol. 30 no. 3, pp. 211-216, 2022.