Financial risk early warning of airlines based on convolutional neural network models

Aviation transportation, as the aerial corridor supporting the global economic operation, has become increasingly significant in the post-pandemic recovery phase. However, beneath the industry prosperity lie numerous risks and challenges. This paper initially elaborates systematically on the rationale for selecting CNN models for conducting research on financial risk early warning, followed by the choice of publicly listed airlines in the A-share market, thereby establishing samples for financial risk early warning and financial health. Subsequently, through differential testing of these two sample categories, suitable financial risk early warning indicators tailored for airlines are scientifically and systematically sifted out. Moreover, to address issues such as the different dimensions of indicator data, the imbalance in the number of sample categories, and dataset partitioning, data preprocessing efforts are undertaken. Finally, the processed data is fed into the CNN model for training, followed by an assessment and analysis of its early warning efficacy.