基于PSO的BP神经网络在腮腺炎发病率预测中的应用  

Application of BP neural network model based on particle swarm optimization for forecasting the incidence of mumps

作　　者：张凡 齐平[2,3] 倪春梅 

机构地区：[1]铜陵市疾病控制中心,安徽铜陵244000 [2]安徽大学计算机学院,安徽合肥230039 [3]铜陵学院数学与计算机科学系,安徽铜陵244000

出　　处：《现代预防医学》2014年第11期1924-1927,共4页Modern Preventive Medicine

基　　金：安徽省自然科学基金项目(1308085QF114);安徽省高等学校省级自然科学研究项目(KJ2013A015)

摘　　要：目的为提供更加准确的预测结果合理调配卫生资源,将基于PSO的BP神经网络应用到铜陵市腮腺炎发病率预测中。方法根据铜陵市2006-2011年的腮腺炎发病率及其相关因素,利用基于PSO的BP神经网络建立腮腺炎发病率预测模型,并对模型的有效性进行验证。结果预测结果显示,PSO-BP神经网络对2012年1-6月铜陵市腮腺炎发病率的预测平均误差为2.57%,预测精度明显高于传统的BP神经网络。结论 PSO-BP神经网络模型能够较好地克服传统BP神经网络易陷入局部极值、收敛速度慢等缺点,为腮腺炎发病率的预测提供了一种有效的方法。Objective To provide more accurate predictions for reasonable distribution of health resource, and the BP neural network model based on particle swarm optimization is applied to forecast the incidence of mumps. Methods According to the incidence of mumps reported by TongLing City form January 2006 to December 2011, the forecasting model based on PSO-BP neural network was constructed, and then the availability of the model was verified. Results From January 2012 to June 2012, the predictive results showed that the mean error of PSO-BP neural network was 2.57%, and the predictive accuracy was significantly higher than the tradition BP neural network. Conclusion PSO-BP neural network model can overcome the limitations of slow convergence and local extreme values for the traditional BP network, and it provides an effective method for forecasting the incidence of mumps.

关 键 词：腮腺炎发病率 预测 BP神经网络 粒子群优化 

分 类 号：R195.1[医药卫生—卫生统计学]