论文部分内容阅读
目的应用自回归移动平均(autoregressive integrated moving average,ARIMA)模型建立宝鸡市细菌性痢疾月发病率预测模型,尝试定量反映宝鸡市辖区内细菌性痢疾流行特征。方法利用2004年1月至2010年12月宝鸡市细菌性痢疾月发病率进行时间序列分析,构建宝鸡市细菌性痢疾月发病率ARIMA模型;利用2004年1月至2010年6月宝鸡市细菌性痢疾月发病率预测2010年7月至2010年12月细菌性痢疾月发病率,并评估预测效果。结果宝鸡市细菌性痢疾月发病率模型为ARIMA(1,0,0)(0,1,1)12,模型自回归参数ARl=0.605(t=6.030,P<0.001),季节移动平均参数SMAl=0.779(t=3.471,P<0.001);残差分析Ljung-Box Q统计量经检验无统计学意义(Ljung-Box Q=8.503,P=0.932)),提示残差为白噪声。建立模型lnYt=0.605lnYt-1+0.799lnYt-12-0.12。2010年7月～12月细菌性痢疾实际月发病率与预测值变动趋势一致,实际值与预测值的绝对误差的绝对值最大为1.896,最小为0.039,相对误差率为6.78%,模型预测效果良好。结论 ARIMA模型可较好地反映宝鸡市细菌性痢疾月发病趋势,可用该模型预测辖区细菌性痢疾发病趋势。
Objective To establish a predictive model of monthly incidence of bacillary dysentery in Baoji City by autoregressive integrated moving average (ARIMA) model and attempt to quantitatively reflect the epidemiological characteristics of bacillary dysentery in Baoji. Methods The monthly incidence of bacillary dysentery in Baoji City from January 2004 to December 2010 was analyzed by time series to construct ARIMA model of monthly incidence of bacterial dysentery in Baoji City. Monthly incidence of diarrhea Predict the monthly incidence of bacterial diarrhea from July 2010 to December 2010 and assess the predictive effect. Results The monthly incidence of bacillary dysentery in Baoji was ARIMA (0,0,0) (0,1,1) 12. The autoregressive parameters ARl = 0.605 (t = 6.030, P <0.001) = 0.779 (t = 3.471, P <0.001); Residual analysis Ljung-Box Q statistic was not statistically significant (Ljung-Box Q = 8.503, P = 0.932)), suggesting that the residual noise is white noise. The model of lnYt = 0.605lnYt-1 + 0.799lnYt-12-0.12. The actual monthly incidence of bacterial dysentery from July to December 2010 has the same trend as the predicted value, and the absolute value of the absolute error of the actual value and the predicted value is the maximum 1.896, the minimum is 0.039, the relative error rate is 6.78%, the model predicts good results. Conclusion The ARIMA model can better reflect the monthly incidence of bacterial dysentery in Baoji City. The model can be used to predict the incidence of bacterial dysentery in Baoji City.