Prediction of Listed Company Growth in Non-public Economy

With the economic reform and open up, the non-public economy has developed rapidly and now has become an indispensable driving force for the national economy development. In the new data era, the governments guide and promote the healthy development of non-public enterprises, which requires the party and the governments accurately identify non-public economic subjects. In this work, an index system including debt paying ability, development ability, operation capability, profitability and tax rate, which aims at the growth ability of non-public economic subjects was established. Kindly check and confirm the edit made in the Data availability statement. Relevant data of non-public enterprises data of A-share listed companies were retrieved from the Wind financial database. The samples were first labeled with factor analysis method and the prediction of enterprise growth was transformed into a supervised classification problem. Then, the Support Vector Machine was further utilized to build models. In the tenfold cross validation, the accuracy (ACC), recall and AUC value were 83.2%, 84.6% and 0.896, respectively. In the independent testing set, the ACC was 82.1% and recall 83.8%, AUC value 0.892, indicating the strong generalization ability of our model. The performance of our model was better than those of other four machine learning algorithms. Finally, the indexes of non-public enterprises were analyzed in this work and found that the profitability and debt paying ability were the most indexes. © 2021, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.