Credit Evaluation for Food Enterprise Based on the Integration of Neural Networks and Expert Ratings

The Food Enterprise Credit Evaluation is of great significance for food supervision and management, for which we have proposed a food enterprise credit evaluation model based on neural networks and expert ratings. This model is to collect data based on the established food enterprise credit evaluation index system, make use of the recurrent neural network structure of the two cycling hidden layers with delayed feedback and an output layer, apply the gradient descent algorithm training neural network of the drive volume and adaptive variable rate to determine the parameters and simulate valuation of the neural network, and then use the Bayesian regularization method to optimize the model of the neural network, and complete determination based on the expert scoring method while comparing and analyzing the network calculation score and optimization results. This model features higher prediction accuracy and stronger versatility for evaluation of the food enterprise credit and boasts important theoretical and practical values for protection of the people's health and promotion of the healthy development of the food industry.