论文部分内容阅读
目的将最小二乘支持向量机(LS-SVM)技术应用到传染病预测中,寻找更加理想的预测结果。方法以某市1991—2002年乙型肝炎(乙肝)月发病率数据建立最小二乘支持向量机预测模型,对2003年1—6月的月发病率进行预测。结果 IS-SVM预测值分别为0.709 9,0.668 1,0.502 5,0.685 1、0.578 5,0.773 7,通过与径向基函数(RBF)神经网络模型和累积式自回归动平均模型(ARIMA)预测结果进行比较,预测精度明显高于RBF网络模型和ARIMA模型,相对误差明显减少,仅为ARIMA模型的23.62%,RBF网络模型的54.69%。结论 LS-SVM模型对乙肝发病率的预测精度更高,效果更好,也验证了支持向量机方法预测能力出色的理论优点,证明了支持向量机技术在传染病预测领域同样有着良好的表现。
Objective To apply least-squares support vector machine (LS-SVM) technique to the prediction of infectious diseases and find out more ideal prediction results. Methods The prediction model of least squares support vector machine was established based on monthly incidence of hepatitis B (hepatitis B) in a city from 1991 to 2002, and the monthly incidence rate from January to June in 2003 was predicted. Results Predicted values of IS-SVM were 0.709 9,0.668 1,0.502 5,0.685 1,0.578 5,0.773 7 respectively. By using Radial Basis Function (RBF) neural network model and cumulative autoregressive moving average model (ARIMA) Compared with RBF network model and ARIMA model, the relative error is obviously reduced, which is only 23.62% of ARIMA model and 54.69% of RBF network model. Conclusion LS-SVM model has higher prediction accuracy and better effect on the incidence of hepatitis B, and also verifies the theoretical advantages of the SVM method. It also proves that the SVM technology also has good performance in the field of infectious disease prediction.