A Predictive Model for Enterprise Development Capability Based on Deep Learning

The prediction of enterprise development capability has become an important part of enterprise management, and traditional enterprise development capability prediction mainly uses subjective evaluation manually. In order to achieve intelligent prediction of industry development capabilities, the author first established an indicator system for predicting the growth of enterprises, including the financial data of listed companies and processed text data such as annual reports and public opinion, so as to predict the growth of enterprises from multiple perspectives. Secondly, using relevant indicators to evaluate the growth of enterprises, providing evaluations of the growth of enterprises in previous years, building a model training set, and evaluated the generalization ability of four classification algorithms on datasets: Random Forest, Adaboost, Support Vector Machine, and Naive Bayes. Then 18 listed companies on the SME board were selected as the research objects, and their growth potential was evaluated using the constructed indicator system and corresponding measurement methods. A growth prediction model was established based on deep learning algorithms. Finally, the growth classification model for small and medium-sized enterprises constructed using deep learning algorithms was tested, the growth evaluation model for small and medium-sized enterprises constructed by the author is more effective, and among the 18 small and medium-sized enterprises, the five enterprises with serial numbers 5, 4, 11, 17, and 2 have high growth potential and investment value. The empirical results indicate that this method can provide more effective reference basis for investment decision-making.