Audit Risk Evaluation Model for Financial Statement Based on Artificial Intelligence

DOI: https://doi.org/10.20532/cit.2020.1005180
2021-07-12
Journal of Computing and Information Technology
Abstract:In recent years, the economy in China has been steadily improving. The financial situation of enterprises in mainstream industries has become the focus of public concern. However, financial statement frauds, which occur frequently, greatly disrupt the economic order in the country. Thus, it is of practical significance to accurately identify and evaluate the audit risks of financial statements. For this purpose, this paper proposes an audit risk evaluation model of financial statement based on artificial neural networks (ANN). Firstly, the authors designed the audit risk indices and quantified the fraud factors of financial statement. Next, an audit risk verification model was established for financial statement and used to verify the predictions on three aspects of financial statement, namely, audit violation penalty (AVP), audit violation announcement (AVA), and financial statement restatement (FSR). Finally, a feedforward neural network was constructed based on the homomorphic encryption algorithm, which was subsequently used to evaluate and predict the audit risks of financial statements. The effectiveness of our model was proved valid through experiments.
What problem does this paper attempt to address?