论文部分内容阅读
[目的]建立中低强度噪声对高频听力损伤关系的BP神经网络模型,探讨BP神经网络模型在职业医学方面的应用。[方法]选择590名中低强度噪声接触者作为研究对象,应用SPSS10.0进行中低强度噪声和高频听力损伤关系的多元线性回归分析,同时应用Matlab7.0建立中低强度噪声和高频听阈关系的BP神经网络模型,比较人工神经网络模型和统计模型的预测效果。[结果]得到由4个输入单元、3个输出单元、18个隐含单元组成的BP神经网络模型,应用BP神经网络预测的误差绝对值平均值和相对误差绝对值都低于应用线性方程的预测结果。[结论]BP神经网络应用于中低强度噪声对高频听力损伤关系的分析效果较好,值得推广应用。
[Objective] To establish a BP neural network model of the relationship between low-intermediate-intensity noise and high-frequency hearing loss and to explore the application of BP neural network model in occupational medicine. [Methods] A total of 590 low-intermediate and low-intensity noise contacts were selected as research objects. SPSS10.0 was used for multiple linear regression analysis of the relationship between low-medium-intensity noise and high-frequency hearing impairment. At the same time, Matlab7.0 was used to establish low- Hearing threshold BP neural network model, the artificial neural network model and the statistical model of the prediction effect. [Result] The BP neural network model composed of 4 input units, 3 output units and 18 implicit units was obtained. The absolute value of absolute error and absolute value of error predicted by BP neural network were both lower than those of linear equations forecast result. [Conclusion] The application of BP neural network in the analysis of the relationship between low and medium intensity hearing loss and high-frequency hearing loss is good and should be popularized and applied.