The Research Of The Credit Rating Of Corporate Environmental Behavior Based On BP Neural Network

In recent years, some enterprises have environmental destruction behavior such as long-term excessive emissions, refusing law enforcement and kangarooing underground pipe. But limited administrative penalties, or difficult to enforce penalties mass incidents have occurred occasionally. Therefore, strengthen the supervision of the corporate environment is still a long-term and urgent task. Corporate environmental behavior of credit evaluation system is studied in this thesis. It can be though the enterprise information and data adopting scientific evaluation method and monitoring means to regulate and deal with the enterprise’s environmental behavior. This will help to solve the problem of enterprise’s illegal sewage, and constraint its environmental behavior effectively, and lay a solid foundation for the country environmental protection enterprise’s sustainable development.In this thesis, firstly, based on the enterprise credit evaluation index, using different indicators of quantitative method; Secondly, combined with the characteristics of the neural network, and build a corporate environmental behavior of credit evaluation model based on neural network; thirdly, on the basis of enterprise environmental behavior of specific applications, an improved adaptive genetic algorithm is proposed, by optimizing the crossover probability and mutation probability, and better solves the premature and slow convergence speed of the genetic algorithm; lastly, put the improved adaptive genetic algorithm applied to the BP neural network system, establish a BP neural network evaluation model based on improved adaptive genetic algorithm, design and implement corporate environmental behavior credit evaluation system.Through simulation experiments show that corporate environmental behavior of credit evaluation model based on neural networks can be very good to make a correct evaluation of corporate environmental behavior. Meanwhile, the improved genetic algorithm can better solve the problem of skilled speed and precocious, so as to search for more accurate data. This algorithm has better robustness. Based on improved BP neural network has higher recognition rate than the BP neural network. It shows that the improved neural network can better learning and evaluate different types of pollution environment of the enterprise credit behavior, to verify the effectiveness and practicability of research results, and provide support for real business work.