The Research On Financial Early Warning Of Listed Companies In China's Manufacturing Industry Based On Multi-layer Peceptron Neural Network

In an environment of rapid economic development and increasingly fierce competition,China's manufacturing industry is constantly evolving under a series of policies,and it faces various problems and risks.In recent years,the situation in which the financial status of listed companies in the manufacturing industry has deteriorated,and the financial crisis has gradually occurred and being special treatment have also occurred from time to time.At present,domestic and foreign scholars have carried out research on the financial crisis warning of listed companies.The listed companies in the manufacturing industry are important and representative in China.Combining the characteristics of the industry,establishing a financial early warning model has become one of the important means of financial warning for manufacturing companies.Therefore,aiming at the actual situation of listed companies in the manufacturing industry,it is necessary to construct a financial early warning model that conforms to the characteristics of the industry,so as to realize the early detection and early resolution of the financial crisis problem,which has far-reaching influence on the sustainable development of the company.By summarizing the existing research results,this paper mainly combines internal control theory,crisis management theory and system non-optimal theory,and it combines theoretical analysis and empirical research and uses the Multi-layer Perceptron Artificial neural Network to study the financial early warning of listed companies in China's manufacturing industry.Firstly,this paper takes China's A-share manufacturing listed companies as the research object,comprehensive domestic scholars' research and the actual situation of China's securities market.This article use the company that is treated by ST(ST sample group)to represent the company's financial crisis,and the company which is not treated by ST.(Non-ST sample group)to represents a good financial position of the company.The sample of this paper is 69 companies that were first processed by ST from 2010 to 2015,and 69 non-ST companies,138 sampleswere selected according to the same industry,the same or similar share capital,and a 1:1 ratio.There are a total of 82 inspection samples from 2016 to 2018.Secondly,using financial indicators and non-financial indicators to construct a financial crisis early warning indicator system,and using K-S test,Independent sample T test,Mann-Whitney U test and principal component analysis,etc.,to screen out the financial crisis and non-financial crisis c.Thirdly,the basic principles of MLP neural network are introduced and the MLP neural network financial early warning model is constructed.Fourth,sample data is selected to verify the validity of the indicator and model.Through the research,the paper draws the following conclusions: 1)The four aspects of the company's debt repayment,operation,profit and development have a significant impact on the financial crisis warning,and the concentration of equity in the non-financial indicators also has a certain impact on the company's financial status;2)The accuracy rate of the ST group and non-ST group financial status prediction results are above 90:%.Generally speaking,this paper combines financial indicators and non-financial indicators,it constructs a financial early warning model for manufacturing industry listed companies based on the MLP neural network,and it determines the financial early warning red line,which has far-reaching impact on risk management and sustainable development of manufacturing enterprises.