摘要
目的比较直线回归模型、对数模型、二次曲线模型、三次曲线模型在肿瘤专科医院出院人数预测拟合效果的优劣,为医院行政部门提供合适的模型。方法应用四种预测方法对肿瘤专科医院出院人数预测并比较拟合值的绝对误差、相对误差和误差平方和。结果对肿瘤专科医院出院人数的直线回归模型的平均绝对误差、平均相对误差绝对值最小为0.72%;其次为二次曲线回归预测模型的绝对误差为337例,相对误差绝对值为1.50%;对数曲线预测模型和三次曲线预测模型的相对误差绝对值比较高,分别为12.06%和12.22%。结论直线回归模型对肿瘤专科医院出院人数短期发展变化规律的分析有比较好的适应性和实用性,可以为肿瘤专科医院今后工作的发展规划提供一定的依据。
Objective We compared the effects on forecasting the number of discharged patients in tumor hospital by the Linear regression, Logarithmic, Quadratic curve, and Cubic regression modds to see which is the best, in order to provide a fitted model for hospital managers. Methods We predicted the number of discharged patients in tumor hospital with the lbur methods mentioned above and compared the absolute error, relative error and the supplier of the fitted values. Result For predicting the number of the discharged patients in the tumor hospital, the value of the mean absolute error, mean relative error of hnear regression curve model is the minimum O. 72% ; while the absolute value of mean error of Quadratic regression is 1. 50% and the absolute value is 337; and those of Logarithmic and cubic regression are higher, 12.06% and 12.22% respectively. Conclusions Linear regression model is one of the best models in predicting the number of discharged patients in tumor hospital, this model can provide more efficient massage about hospital development.
出处
《中国病案》
2008年第8期46-48,共3页
Chinese Medical Record
