| The 2021 Government Work Report proposed to accelerate the digital development,create new advantages of the digital economy,and promote the digital transformation of the industry.In the 21st century,digital enterprises will flourish everywhere,which is an inevitable trend.Especially after the impact of the COVID-19 epidemic on the economy,the digital economy is becoming more and more popular in all industries.In order to make digitalization develop as rapidly as possible in various industries,we must pay attention to the growth and development ability of digital enterprise clusters.One of the important characteristics of digital enterprises is growth.It is not only the basis of investors’ investment,but also an important indicator monitored by national regulatory agencies.At the same time,enterprise growth is also crucial to the development and prosperity of the national economy.Therefore,analyzing enterprise growth is the key to measuring the current situation and future development of enterprises.At present,the comprehensive prediction research on the growth of industrial digital enterprises is still relatively small.Many scholars have not built a unique growth prediction index system and prediction model based on the characteristics of industrial digital enterprises.Therefore,it is of great significance to study the comprehensive growth prediction model of industrial digital enterprises for enriching the research on the growth of digital enterprises.Based on the analysis of the research status of the growth of industrial digital enterprises at home and abroad,this study selected 1072 industrial digital listed companies as the research objects,and the sample period was selected from 2017 to 2021,with certain timeliness.Considering the internal factors,external factors,financial factors and non-financial factors that affect the growth and development of enterprises,the growth prediction index system based on financial data and the growth prediction index system based on financial data and annual report text are respectively constructed to comprehensively evaluate and predict the development and growth of enterprises.Since this study has built a machine learning classification prediction model and a deep learning classification prediction model,three growth explicit result indicators,namely,net profit growth rate,main business income growth rate and total asset growth rate,are selected,and the growth comprehensive result indicator is obtained by entropy method as the training label of the model.Based on the selected sample data,this study constructs SVM model,random forest model,XGBoost model,LightGBM model and convolution neural network model,and selects the best performance model through horizontal comparison of AUC values.Through comprehensive comparison,it is found that whether based on financial data or financial data and annual report text data,the classification and prediction ability of the convolution neural network is the best,and after the introduction of the annual report text index data,the prediction ability of the convolution neural network is futher improved,so this paper selects the convolution neural network model based on financial data and annual report text as the comprehensive model for the growth prediction of industrial digital enterprises.To some extent,this study enriches the current research on the development of industrial digitalization and the growth prediction of industrial digitalized enterprises and the application of in-depth learning model,and can also provide some basic reference for further analysis and research in the future. |
