Research On Financial Risk Early Warning Of Communication Equipment Manufacturing Industry Based On Neural Network Model

In 2019,under the guidance of the spirit of the 19 th National Congress of the Communist Party of China,my country embarked on a new journey of building a modern socialist country in an all-round way.Under the guidance of national macroeconomic policies,my country's communications equipment manufacturing industry has also achieved rapid development,and a number of top communications equipment manufacturing companies such as ZTE and Huawei have emerged.At the same time,as the pace of global economic integration accelerates,the competition faced by enterprises is becoming increasingly fierce.The communication equipment manufacturing industry is also facing huge operating risks.During the development process,it has exposed the incomplete financial risk early warning system in the industry and financial risks.Problems such as backward management have become important factors restricting the further development of such enterprises.Therefore,it is urgent to build a financial risk early warning model for the communication equipment manufacturing industry.This article takes the communication equipment manufacturing industry as the research object,uses literature research methods,quantitative analysis methods,and case analysis methods to analyze the financial r isks of the industry,constructs an early warning indicator system suitable for the industry,and establishes a suitable system based on mathematical modeling.The financial risk early warning model of the communication equipment manufacturing industry is expected to improve the financial risk early warning research of the communication equipment manufacturing industry.Based on reading a large number of relevant documents,this article first sorts out the research on financial risk early warning definition,mechanism,warning situation division and early warning model.Secondly,it introduces the definition,characteristics,functions and processes of financial risk early warning,introduces the classic financial risk early warning model,elaborates the construction process of financial risk early warning model,and clarifies the theoretical basis of financial risk early warning.Next,it analyzes the industry characteristics of the communication equipment manufacturing industry and the characteristics of the financial risks faced by the industry.Build a financial risk early warning model for the communication equipment manufacturing industry based on the neural network model again.The first step selected 22 dimensions of financial and non-financial index d ata of 86 listed companies in the communications equipment manufacturing industry as the research sample;the second step used the K-means clustering algorithm to determine the three types of financial risks faced by the company;third Step adopts random down-sampling and SMOTE up-sampling methods to deal with unbalanced samples;the fourth step establishes a neural network early warning model.Then use the built neural network model to predict the financial risk status of D company,and find that D company currently faces a higher overall financial risk.Then we put forward different solutions for each type of risk: for debt repayment risks,we need to broaden financing channels and optimize capital structure;for profit risks,we need to accelerate product transformation and upgrading to improve profitability;for operational risks,we need to strengthen Accounts receivable management and improvement of inventory management level;for innovation risks,it is necessary to increase R&D investment;for litigation risks,it is necessary to strengthen internal management.The last part is the conclusion and recommendations of this article.This article draws the following conclusions: First,the neural network early-warning model constructed in this article can effectively provide financial risk early warning for the communication equipment manufacturing industry one year in advance.The neural network model constructed in this article with a 19-8-3 network structure is based on the training set.The accuracy rate is as high as 95%,and the accuracy rate on the test set is 87.5%,indicating that the model has a good prediction effect.Second,the selection of model parameters affects the accuracy of early warning.The use of different hidden layer nodes,optimization algorithms,loss functions and activation functions will lead to certain differences in the accuracy of early warnings.The third is that balancing the sample data helps the neural network model to learn more features,thereby improving the accuracy of the early warning model.In order to better apply the model,this article proposes the following three application suggestions: O ne is to expand the sample data volume and the dimensions of early warning indicators,for example,sample data of non-listed companies and more non-financial indicators can be included To the early warning system;the second is to refine the classification of the company's financial risk level,for example,the clustering results can be further subdivided into five categories: healthy,good,general,light and serious;third,to adapt to the development of the times and constantly update the nerves Network early warning models,for example,deep learning models such as recurrent neural network(RNN)and convolutional neural network(CNN)can be used as financial risk early warning models to improve the application effect of the early warning model.