【摘要】 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。更多还原

【Abstract】 [Objective] To use least squares support vector machine(LS-SVM) technology in infectious disease forecasting,in order to find a more perfect prediction results.[Methods]The forecasting model based on least squares support vector machine was constructed under data from hepatitis B monthly incidence rate reports in a city from 1991-2002,and then the incidence rates form Jan.2003-Jun.2003 was forecasted by the established model.[Results] The incidence of LS-SVM predictive values were 0.709 9,0.668 1,0.502 5,0.685 1,0.578 5,0.773 7,comparing with RBF network model and ARIMA model,the prediction accuracy was significantly higher than that of RBF model and ARIMA model,the relative error was significantly reduced,only 23.62% of the ARIMA model and 54.69% of the RBF network model.[Conclusion]LS-SVM model is more effective and accurate for predicting the incidence rates of infectious disease,which also verifies the superiority of support vector machine theory used in forecasting and proved that the support vector machine technology in the field of infectious disease forecasting the same good performance.更多还原