论文部分内容阅读
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 To compare the advantages and disadvantages of the linear regression model, logarithmic model, quadratic curve model and cubic curve model in the prediction of the number of discharged people in oncology hospitals, and provide appropriate models for hospital administrative departments. Methods Four prediction methods were used to predict the number of discharged people in the cancer specialist hospital and to compare the absolute error, relative error and sum of squared errors of the fitted values. Results The absolute value of absolute absolute error and average relative error of the linear regression model for the number of patients discharged from the cancer hospital were 0.72%. The absolute error of the quadratic curve regression model was 337, and the absolute value of the relative error was 1.50%. The absolute values of the relative error between the number curve prediction model and the cubic curve prediction model are relatively high, which are 12.06% and 12.22%, respectively. Conclusion The linear regression model has better adaptability and practicability for the analysis of the short-term development and changes of the number of discharged people in the oncology hospital. It can provide a basis for the future development planning of the oncology hospital.