论文部分内容阅读
目的探讨构建并应用自回归求和移动平均(autoregressive integrated moving average,ARIMA)模型预测原静安区成人流感样病例(influenza-like illness,ILI)就诊百分比的可行性。方法基于2011—2014年上海市原静安区的逐月成人ILI就诊百分比,模型参数确定采用非条件最小二乘法,模型结构依据简洁与残差不相关原则确定,拟合优度以许瓦兹贝叶斯准则与赤池信息准则评估,构建成人ILI就诊百分比预测的最优ARIMA模型。以模型预测原静安区2015年1—10月成人ILI就诊百分比,计算实际值与预测值的相对误差;并预测原静安区2016年的成人ILI就诊百分比。结果模型ARIMA(0,2,1)(1,1,0)12(无常数项)对成人ILI就诊百分比时间序列拟合良好,移动平均参数(MA1=0.944)与季节自回归参数(SAR1=-0.542)有统计学意义(P<0.001),残差达到白噪声(P>0.05),模型表达式为(1+0.542B)(1-B)~2(1-B~(12))Zt=(1-0.944B)μt。2015年1—10月的成人ILI就诊百分比的预测值符合实际值的变动趋势,相对误差最小仅为4.45%。结论 ARIMA模型可以较好地拟合原静安区成人ILI就诊百分比的时间变动趋势,能对成人ILI就诊百分比进行预测,短期预测有较高的精度。
Objective To investigate the feasibility of constructing and applying the ARIMA model to predict the percentage of treatment-oriented patients with influenza-like illness (ILI) in Jingan district. Methods Based on the percentages of monthly ILI visits from 2011 to 2014 in Shanghai Jing’an District, the model parameters were determined by using non-conditional least-squares method. The model structure was determined according to the principle of concision and unrelated residuals. The goodness-of-fit Sri Lanka and Chi Chi information criteria assessment, to establish the optimal ARIMA model for the prediction of the percentage of adult ILI treatment. The model was used to predict the percentage of adult ILI visits from January to October 2015 in Jingan district and calculate the relative error between the actual value and the predicted value. And the percentage of adult ILI in original Jingan District in 2016 was predicted. Results The model ARIMA (0,2,1) (1,1,0) 12 (no constant) fitted well to the percentage of adult ILI visit percentages with the moving average (MA1 = 0.944) and the seasonal autoregressive (SAR1 = -0.542), and the residuals reached white noise (P> 0.05). The expression of the model was (1 + 0.542B) (1-B) ~ 2 Zt = (1-0.944B) μt. Estimates of the percentage of adult ILI visits from January to October 2015 are in line with the trend of actual value with a relative error of at least 4.45%. Conclusions The ARIMA model can well fit the trend of the time percentage change of adult ILI in adult Jingan district and predict the percentage of adult ILI treatment. The short-term prediction has higher accuracy.