Research On Financial Distress Early Warning Model Based On Quantum Bird Swarm Algorithm-Artificial Neural Network

Listed companies are an essential part of China's national economy.With the development of the market economy,the competition between enterprises is increasingly fierce,especially the external financial factors such as Sino-US trade war,the economic downturn,and other factors,the financial risks faced by listed companies.And the possibility of a bankruptcy crisis has risen sharply,and the demand for financial warnings by listed companies has become more urgent.The deterioration of the profitability of listed companies will not only threaten the interests of the company itself but also expose investors to significant financial losses.In recent years,there have been many listed companies in China's A-share market that have been specially dealt with because of financial problems.Due to the current asymmetric information in China's securities market,the financial crisis of enterprises may even harm the market environment.Therefore,whether from the perspective of the enterprise itself or the view of stakeholders,how to establish an ongoing financial crisis early warning mechanism is a critical research topic.Early warning of the financial distress,the main content is to monitor the economic situation in real time,through the analysis and mining of data,to obtain useful information to predict the financial condition of the enterprise.At present,there are many kinds of financial crisis warning models in the academic world,such as traditional linear discriminant analysis,multiple regression,logistic regression,and current BP models and SVM models.To obtain higher prediction accuracy,techniques such as statistical methods and artificial intelligence methods are continuously optimized for models.In the financial early warning model,the selection of early warning indicators is directly related to the performance of the early warning model.In the existing financial early warning model,BP artificial neural network shows excellent advantages.Its basic working principle is through the training of samples,which is the network with similar human brain learning,memory and recognition ability-information processing function.The self-learning ability and self-association ability of BP artificial neural network are more prominent.When the sample data is missing,and the parameters may drift,the relatively stable output can also be given,so it is very suitable for establishing the financial crisis early warning model.However,at present,the selection of early warning indicators based on most models of BP neural network is still difficult to avoid personal subjectivity,or the model selects too many symbols,the operation is complicated,and over-fitting may occur,or the chosen indicators are insufficient,and symbols may exist relevance between them,affecting the accuracy of the model.The bird swarm algorithm is a new group intelligent optimization algorithm.The basic idea of BSA is to organize individual birds according to specific rules.Each's behavior affects the whole bird group,and the evolution of the population is realized through the continually updated position of the bird group.The bird swarm algorithm can effectively solve complex multidimensional optimization problems and many problems.However,the existing bird swarm algorithm can only address the continuous optimization problem and cannot be applied to the discrete problem.The financial early warning problem studied in this paper is a separate problem;that is,the existing bird swarm algorithm cannot be directly applied to the financial warning problem.This paper proposes a quantum bird swarm algorithm combining bird feeding mechanism and quantum computing.It can effectively solve the discrete optimization problem,especially when the influencing factors are more;the advantage of fast screening will be revealed.It can be used in effective detection of financial early warning indicators.Therefore,this paper combines quantum bird swarm algorithm and BP artificial neural network,comprehensively uses financial crisis dynamic early warning theory,intelligent algorithm,MATLAB programming technology,through data mining,establishes QBSA-BP based neural network from time series and cross-section.The early warning model:the quantum bird swarm algorithm is used to select the financial distress early warning indicators,which are input into the BP artificial neural network for prediction.The dynamic early warning theory and method of the enterprise financial distress are systematically studied and adopted,and a large sample is adopted.The multivariate,long-term sequence data was empirically studied,and the early warning model showed good robustness and predictive ability.Through data mining,this paper selects 80 enterprises in the A-share listed companies of the Shanghai and Shenzhen Stock Exchanges for the first time in 2016-2018,and pairs 160 non-STs according to the principle of the same industry,the same period,and similar asset scale.Enterprises,a total of 240 companies constitute a sample,using the quantum bird swarm algorithm to select the index input BP artificial neural network to establish a micro-level financial distress dynamic early warning indicator system based on listed companies,focusing on solving the theory of creating financial distress early warning indicator system basis and integrity issues.The research results show that the financial distress early warning model based on quantum bird swarm algorithm-BP artificial neural network has good predictive ability.The empirical research on 240 sample companies shows that the method shows superiority in error rate and recognition accuracy.The performance,which is a huge improvement for the financial crisis early warning model,further expands the in-depth application of artificial intelligence methods in the field of financial crisis early warning,with broadly application foreground.