www.brydzyk.fora.pl

defgf96tj

Low-intensity noise in high-frequency hearing loss caused by BP neural network model

Threshold as 2 Comparison of the prediction models to calculate the error of absolute and relative absolute error: The results of BP neural network prediction mean absolute error and relative absolute error lower than predictions of linear equations, BP Neural network prediction better. The results in Table 1. Table l4000HZ hearing threshold effect as comparing the two models predict 3 to discuss artificial neural network (ArtificialNeuralNetwork, ANN) is developed rapidly in recent years, a set of neuroscience, computer science, information science, engineering science as one of the edge of the cross-disciplinary, has a unique way of information storage,[link widoczny dla zalogowanych], good fault tolerance, large-scale parallel processing and a strong self-organization, self-learning and adaptive capacity,[link widoczny dla zalogowanych], has been used in signal processing,[link widoczny dla zalogowanych], pattern recognition, forecasting and other fields, and has a wide range of applications prospects. BP network (BackpropagationNet-work) is a development in theory and applications are more mature back-propagation method for multi-layer network. BP neural network model will be applied to the low intensity noise on high-frequency hearing loss study of the effects, its biggest advantage is that the design does not require any mathematical model of a single mathematical formula can overcome the mechanical and statistical principles in the combined sample independent division, improve training a neural network model robustness, avoiding the direct use of linear regression model over-fitting problem, with more flexibility, but by the interaction between neurons to complete the entire network information processing, which leads to better prediction results, and can be more than one different output results. With this other on the BP neural network model applied in the medical field was similar to the literature J, results suggest that: the relative multiple regression equation,[link widoczny dla zalogowanych], BP neural network model has many obvious advantages, the BP neural network model was applied to high intensity noise on high frequency hearing injury forecast is simple and feasible and effective prediction of occupational diseases in the future will be more widely used in the field,[link widoczny dla zalogowanych], in order to provide a powerful tool for the prevention and treatment of occupational diseases. Application of BP neural network model of high intensity noise on the dynamic high-frequency hearing loss prediction, is the artificial neural network applications in the field of occupational medicine exploration results in this paper show that: it can improve the prediction of the effect of the statistical model than the traditional a distinct advantage. And as the greater the number of training samples, the higher the accuracy of the network, once trained, a good network model can be used to predict the actual work proposed corresponding protection measures.
More articles related to topics：

[link widoczny dla zalogowanych]

[link widoczny dla zalogowanych]

[link widoczny dla zalogowanych]