疟疾发病人数灰色预测模型中参数稳定性的探讨  

STUDY ON THE STABILITY OF PARAMETERS IN GRY FORECASTING MODELS OF MALARIA INCIDENCE

作　　者：方莹[1] 汤林华[1] 顾政诚[1] 

机构地区：[1]中国预防医学科学院寄生虫病研究所

出　　处：《中国寄生虫学与寄生虫病杂志》1996年第2期115-118,共4页Chinese Journal of Parasitology and Parasitic Diseases

摘　　要：目的：探讨建立参数稳定的疟疾发病人数灰色预测模型的方法。方法：用经病例增长率处理数据的ＲＣＩ．ＧＭ（１，１）法建模来取代直接的ＧＭ（１，１）法建模。结果：ＲＣＩ．ＧＭ（１，１）建模法改善了模型稳定性，即ＲＣＩ．ＧＭ（１，１）的模型值不会因数据稍有波动而受严重影响。同时也减少建模误差率和预测误差率。结论：ＲＣＩ．ＧＭ（１，１）法建模稳定性好，因此能提高预测模型的实用意义。AIM: To establish a forecasting model of GM(1,1) with stable parameters for improving the prediction of malaria incidence. METHOD: Using the RCI.GM(1,1)[i.e.modelling GM(1,1) with rate of case increase] method to replace the direct GM(1,1) method. RESULTS: The RCI.GM(1,1) model increases the stability of parameters and reduces the modelling and forecasting error rates. The output of RCI.GM(1,1) model would not be influenced seriously by slight fluctuation of data.CONCLUSION: The RCI.GM(1,1) model is better than that of GM(1,1) for improving the modelling and predictive value of malaria forecasting model.

关 键 词：疟疾 预测模型 稳定性 

分 类 号：R531.304[医药卫生—内科学]