Analysis And Prediction Of Factors Affecting The Performance Of Scientific And Technological Innovation Enterprises Based On Panel Model And BP Neural Networ

With the continuous development of our market economy,the competition among economic subjects in knowledge,technology and innovation has become the main competition pattern,and innovation has become the main way for enterprises,industries and even countries to build sustainable competitive advantages.Innovation is not only a means for an enterprise to have competitiveness,but also the purpose of its development.As the core competence of an enterprise,innovation can enhance its value,accelerate its growth and increase its income.Therefore,it is of great significance to objectively and accurately analyze the important factors affecting the performance of sci-tech innovation-oriented enterprises and predict their performance based on this for promoting the development of sci-tech innovation-oriented enterprises and enhancing their competitive advantages.In order to study the important factors affecting the performance of science and technology innovative enterprises and make a scientific and reasonable forecast of the development trend of enterprise performance in the future,In this paper,321 listed enterprises in computer communication and other electronic equipment manufacturing,pharmaceutical manufacturing,information transmission,software and information technology service industry in Shanghai and Shenzhen main board A-shares and growth enterprise Board from2011 to 2020 are taken as research samples,combined with literature review method,comparative analysis method,panel regression analysis,BP neural network and other empirical analysis methods.Firstly,according to the existing relevant literature and theories,the important factors that may affect the performance of science and technology innovative enterprises are summarized and sorted out,and these important factors are studied and analyzed,and the corresponding research hypothesis is made.Then,the panel fixed effect model is established from the micro and macro perspectives to analyze the important factors affecting the performance of sci-tech innovative enterprises.Then,based on the results of empirical analysis,the important factors affecting the performance of science and technology innovative enterprises are screened and determined,and the single index and comprehensive index BP neural network prediction models are established respectively to predict the performance of science and technology innovative enterprises.Finally,Bayesian regularization algorithm(BR)and genetic algorithm(GA)were used to optimize the original BP neural network prediction model of comprehensive index respectively.Compared with the difference of model prediction performance before and after optimization,the optimal model was selected to predict the performance of science and technology innovative enterprises.The main conclusions of this paper are as follows:(1)The empirical analysis results from the micro(internal)perspective show that the asset structure,debt paying ability,operation ability,research and development ability,innovation ability and ownership structure of science and technology innovation-oriented enterprises have a significant impact on enterprise performance.(2)The empirical analysis results from the macro(external)perspective show that,Macroeconomic growth,inflation,the degree of diversification of enterprises,prudent and appropriately loose monetary policy and fiscal policy have significant effects on enterprise performance.(3)The prediction results show that the prediction performance of the comprehensive index BP neural network model is significantly higher than that of the single index BP neural network model.The prediction performance of GA-BP neural network model is obviously better than that of BR-BP neural network model and original model.