论文部分内容阅读
目的探讨应用自回归滑动平均混合模型(Autoregressive Integrated Moving Average Model,ARI MA)预测甘肃省麻疹发病率的可行性。方法应用SPSS13.0软件对甘肃省1995~2006年麻疹逐月发病率进行ARI MA建模拟合;按照残差不相关原则、简洁原则、赤迟准则与贝叶斯准则建立ARI MA麻疹预测模型,用2007年分月发病数检验模型。结果模型ARI MA(1,0,0)(0,1,1)12(不含常数项)所有参数都通过统计学检验,残差序列是白噪声,拟合优度相对最好,参数间也无明显相关性(r=0.069)。结论ARI MA模型很好地模拟和预测了麻疹既往发病周期性以及各月发病数,将其应用于甘肃省麻疹发病预测是可行的。
Objective To explore the feasibility of predicting the incidence of measles in Gansu Province by Autoregressive Integrated Moving Average Model (ARI MA). Methods SPSS 13.0 was used to simulate the monthly incidence of measles in Gansu Province from January to December. ARI MA measles prediction model was established according to the principle of residual unrelated, simple principle, late-retarded criterion and Bayesian criterion , With the monthly incidence of 2007 test model. Results The model ARI MA (1,0,0) (0,1,1) 12 (excluding the constant term) All parameters were tested statistically, the residual sequence was white noise, the best goodness of fit, between parameters There was also no significant correlation (r = 0.069). Conclusion The ARI MA model can simulate and predict the periodic incidence of measles and the monthly incidence of measles. It is feasible to apply it to the prediction of measles in Gansu Province.