Financial performance evaluation and forecasting of enterprises by the combination of PCA and CNN deep learning

1 Introduction Financial evaluation focuses on pursuing sustainable financial development under the“New Normal”of China’s economy.It is of great practical significance to explore innovative methods based on established financial evaluation indexes and systems[1,2].The joint strategy of PCA methods and Deep Learning models seems to have a scientific basis and a good fusion mechanism in earlier studies[3-5].The innovation of our research is to combine evaluation method,unsupervised learning,with supervised learning.Based on the PCA method,the selected data is dimensionally reduced and comprehensively scored,then the unsupervised learning K-means clustering method is used to divide the sample into five evaluation levels according to the performance score.Finally,combined with the Convolutional Neural Network(CNN)framework,the company’s financial performance can be profiled and predicted.On one hand,the theoretical significance of this study is to combine the traditional PCA method with the Deep Learning method to predict the performance level for the performance portrait of enterprises.