Research on the Construction of Enterprise Financial Fraud Identification Model under Data Mining Technology

Under the influence of digital economy, the rapid development of capital markets at all levels has not only promoted technological innovation, product innovation and market structure changes, but also prompted new risks and challenges to show a trend of continuous increase, especially the frequent exposure of enterprise financial fraud incidents, which not only seriously damaged the interests of investors, but also undermined the fairness and transparency of the market. Faced with this situation, the traditional financial audit methods often rely on manual audit and sampling inspection, which has obvious limitations and can no longer meet the practical application requirements. In view of the current problems, this paper will discuss the application feasibility of data mining technology in this field, and put forward a set of hybrid enterprise financial fraud identification model construction scheme to make up for the shortcomings of traditional audit methods and improve the accuracy of identification results. Practice has proved that the whole model contains two parts: a primary learner and a secondary learner, in which the primary learning is composed of SVM model and the secondary learner is composed of Logistic regression model, in order to improve the accuracy and stability of the recognition results and provide strong technical support and guarantee for the healthy development of the capital market.