论文部分内容阅读
目的应用GM(1,1)、二次趋势和指数趋势模型对我国梅毒发病率进行预测,比较其预测效果。方法应用三种模型对我国2006-2012年梅毒年发病率数据进行拟合建模并选取最优模型进行外推预测。结果三种模型的MAPE均未超过10%,预测精度较好,其中二次趋势方程的MAPE为2.39%,拟合精度最佳,外推预测2013年、2014年梅毒发病率分别为32.90/10万、34.40/10万。结论二次趋势方程对我国2006-2012年梅毒的发病率数据拟合结果最理想,可用于梅毒发病率的短期预测。
Objective To forecast the incidence of syphilis in China using GM (1,1), quadratic trend and exponential trend model and compare the predictive value. Methods Three models were used to model the annual incidence of syphilis in China from 2006 to 2012, and the optimal model was selected for extrapolation prediction. Results The MAPE of the three models did not exceed 10%, and the prediction accuracy was good. The MAPE of the second trend equation was 2.39%, and the fitting accuracy was the best. The extrapolation predicted that the incidence rates of syphilis in 2013 and 2014 were 32.90 / 10 Million, 34.40 / 100,000. Conclusions The quadratic trend equation has the best fitting result for the incidence of syphilis in China from 2006 to 2012 and can be used for the short-term prediction of the incidence of syphilis.