【摘要】 目的 :应用灰色模型预测克拉玛依市肺结核发病率 ,为合理调配结核病防治的卫生资源提供科学依据。方法 :建立灰色预测模型 ,并与线性回归模型、指数模型、多项式模型拟合效果进行比较。结果 :建立的灰色预测模型α为 - 0 . 0 78,u为 4 7.899,其预测模型精度指标后验差比值 C为 0 .390 ,小误差频率 P为 1.0 0 ,属于“合格”一级。预测 2 0 0 0年克拉玛依市肺结核发病率为 72 .4 1/ 10万 ,相对误差为 - 8.6 1 ,较其余三种模型预测精度高。结论 :GM (1,1)模型可以对该地区肺结核发病率做较好的短期预测更多还原

【Abstract】 Objective:To provide the best model for predicting the incidence of Tuberculosis in Kelamayi and scientific base for reasonable distribution of health resource of Tuberculosis prevention and treatment.Methods:(1)To establish GM(1,1), according to the incidence of Tuberculosis in Kelamayi from 1995 to 1999, and then do Curve fit by use of Linear Regression model, Exponential model and Polynomial model.(2)To forecast the incidence of Tuberculosis in 2000 in terms of four models.(3)To contrast the fitness and the prognosticating precision of four models.Results:(1)The prognosticating precision of GM(1,1) ( C=0 390,P=1 00 ) was eligible in accordance with the criterion of Grey model prognostication assessment. α and u ,the coefficients of the Grey model, were -0 078,47 899,respectively. (2)As far as the fitness, the determination coefficient (R 2) of GM(1,1) (R 2=0 886) was higher than that of Linear regression model (R 2=0 622), Exponential model (R 2=0 670) and Polynomial model (R 2=0 764). (3)According to the Grey model, the incidences of Tuberculosis in 2000 estimated in light of GM(1,1), Linear Regression model, Exponential model and Polynomial model were 72 41, 75 45, 78 99 and 32 28 per 10 thousand population,respectively. The prognosticating relative error of GM(1,1) (-8 61%) was the least among all the models.Conclusion:GM(1,1) is more appropriate than other models in short term prediction of the incidence of Tuberculosis in Kelamayi.更多还原