Research on Comprehensive Financial Crisis Warning Model of Listed Companies Based on Financial Report Big Data and Deep Learning Technology

Against the backdrop of an increasingly complex global economy, corporate financial risks continue to rise, prompting the establishment of an efficient risk warning model to become a focus of academic and practical attention. With the widespread application of artificial intelligence technology, deep learning has opened up a new path in the identification and assessment of financial risks. Based on financial statement big data, this paper applies mainstream deep learning algorithms to construct a comprehensive financial crisis warning model for listed companies. The study selected data from listed manufacturing companies and used the propensity score matching method (PSM) to divide the samples into two groups: financial crisis and stable operation. A multi-level and multi-dimensional warning indicator system was established, and the accuracy of the model was verified by computer simulation and multi-model comparison. The results showed that the prediction accuracy of the model based on deep learning exceeded 90%. This paper uses Logistic regression analysis to identify important factors affecting financial risks. This study not only enriches the financial risk warning framework in theory, but also provides useful tools and references for managers to improve financial security management.