Research on big data-driven accounting anomaly detection and intelligent audit model based on deep learning algorithm

As big data technology progresses swiftly, the volume and complexity of accounting data are escalating. Traditional anomaly detection and auditing methods struggle with low efficiency and accuracy when confronted with vast amounts of unstructured financial data. Meanwhile, with the rapid evolution of artificial intelligence and big data, companies are actively exploring and developing a financial shared service model, which includes the establishment of an intelligent audit platform. Deep learning algorithms show a broad application prospect in the field of accounting anomaly detection and intelligent auditing by virtue of their powerful feature extraction and pattern recognition capabilities. A big data-driven accounting anomaly detection and intelligent audit model is constructed. The needs of accounting anomaly detection in big data environment are analyzed, and the feasibility of deep learning in this field is explored. Design a deep learning-based accounting anomaly detection model, including data preprocessing, feature engineering, deep neural network architecture optimization and model training and evaluation. Propose a framework of intelligent auditing system to realize the intelligence of financial text auditing and auditing process by combining natural language processing (NLP) and other technologies. The model outperforms traditional methods in terms of accuracy of anomaly detection and auditing efficiency. The findings of the study offer significant validation for the advancement of intelligent auditing technology and serve as a valuable benchmark for subsequent research into AI-based financial auditing.