Abstract:This study investigated the accuracy of back propagation artificial neural network model for prediction of the growth of Staphylococcus aureus in sauced beef. The data of Staphylococcus aureus growing in sauced beef were collected at different temperatures (15, 25, 36 ℃) with different initial inoculations (10< sup> 2< /sup>,10< sup> 3 sup>,10< sup> 4< /sup> CFU/mL). With the saftwares of Python and MATLAB, the network structure with the best fitting effect was screened out and the back propagation neural network model was constructed. At the same time, the modified Gompertz model was established for the comparison. The accuracy of the two models was tested using square root error, bias and accuracy factors. The results showed that the model with the network structure of 2-35-30 had the best fitting. The errors generated in the back propagation neural network model and the modified Gompertz model were within an acceptable range. Compared with that of the modified Gompertz model, the error of back propagation neural network was smaller and it could be used to predict the growth of Staphylococcus aureus in sauced beef with higher accuracy.