基于变系数回归的高校大学生艾滋病新增病例的预测研究

Forecast of the number of new AIDS reports for college students based on the variable coefficient regression model

利用2007—2016年湖南省高校大学生艾滋病新报告病例的数据,将2011—2016年艾滋病发病数据作为检验样本,以平均相对误差作为评价指标,建立变系数回归模型,并对2017年新报告病例数进行预测。经计算得到变系数回归模型的平均相对误差为0.6801%,灰色预测方法的平均相对误差达到20.9792%,以及2017年湖南省高校大学生艾滋病新报告病例数的预测区间为[209, 231]。结果表明变系数回归模型预测效果好于灰色预测,更适用于湖南省高校大学生艾滋病疫情预测。

A regression model of variable coefficient is built based on the new AIDS report cases of university students in Hunan Province from 2007 to 2016, while the average relative error is participate as the evaluation index,the 2011—2016 AIDS incidence data are regarded as the test sample to make a forecast for the number of new reported cases in 2017. The average relative error of the variable coefficient regression model is 0.6801%, the average relative error of the grey prediction method is 20.9792%, and the prediction interval of new report cases of university students in Hunan province in 2017 is (209, 231)The results show that the prediction effect of the variable coefficient regression model is better than the grey prediction. It is more suitable for the prediction of the AIDS situation of college students in Hunan province.