论文部分内容阅读
[目的]运用灰色模型预测某市肺结核发病率。探索数据波动较大的情况下,适用的灰色模型预测方法。[方法]根据某市1997~2008年肺结核发病率数据分别建立动态GM(1,1),新陈代谢GM(1,1)和常规GM(1,1)。比较3种模型预测精度和准确性,选择合适肺结核发病率的预测模型并进行外推预测。[结果]检验性预测中,3种模型预测精度均为一级,预测值相对误差依次为-13.56%、-12.27%和-31.18%。将动态GM(1,1)和新陈代谢GM(1,1)作为该市肺结核发病率预测模型,两种预测模型的预测精度分别为二级和四级。采用动态GM(1,1)对该市肺结核发病率进行外推预测,2009年、2010年该市肺结核发病率分别为90.22/10万、89.66/10万。[结论]针对该市肺结核发病率数据波动较大的情况,采用动态GM(1,1)进行预测是比较适用的方法。
[Objective] To predict the incidence of tuberculosis in a city by using gray model. Explore the case of large data fluctuations, the application of the gray model prediction method. [Method] According to the incidence of tuberculosis in a city from 1997 to 2008, the dynamic GM (1,1), metabolism GM (1,1) and routine GM (1,1) were established respectively. The prediction accuracy and accuracy of the three models were compared. The prediction model of the incidence of pulmonary tuberculosis was selected and extrapolated. [Results] The predictive accuracy of the three models were all in the first grade, and the relative errors of the predicted values were -13.56%, -12.27% and -31.18%, respectively. Taking dynamic GM (1,1) and metabolic GM (1,1) as the prediction model of pulmonary tuberculosis incidence in the city, the prediction accuracy of the two prediction models are respectively Grade 2 and Grade 4. The dynamic GM (1,1) was used to extrapolate the incidence of tuberculosis in the city. The incidence rates of tuberculosis in this city in 2009 and 2010 were respectively 90.22 / 100000 and 89.66 / 100000. [Conclusion] The prediction of dynamic GM (1,1) is a suitable method for the case that the incidence of tuberculosis in the city fluctuates greatly.