Classification method for safety status of industrial enterprise based on BP neural network

Abstract

For the purpose of exploring the classification method for safety status of industrial enterprise, a safety index system was constructed and a model based on back propagation (BP) neural network was also built and applied. The safety index system includes 5 factors and 20 properties. Industrial enterprise can be classified into 4 levels: D, C, B and A, where A is the highest level. 14 training samples and 6 test samples was chosen to train and test the BP neural network until the results achieved the high accuracy, and then the model was used in practical application. The results show that the safety index system is comprehensive and scientific, and the model based on BP neural network is effective enough to reasonably classify safety status of industrial enterprise. This method can be helpful for safety management of government and enterprises themselves.