论文部分内容阅读
目的建立基于BP神经网络的住院天数拟合模型,并在已建立的神经网络模型的基础上,进行住院天数的预测和影响因素的敏感度分析,利用本研究的建模结果,为BP神经网络建模的方法学提供一定的参考依据,并能帮助卫生管理决策者做出正确的决策和分析。方法利用SQL提取HIS数据,在Clementine 11.1中进行建模和预测,预测结果用SPSS16.0进行假设检验。结果BP神经网络的拟合度和预测准确度分别为96.678%和86.67%,术前住院天数对射频消融术患者的住院天数影响最大。结论BP神经网络相对其他传统统计方法而言,是比较适合于住院天数数据特征的建模方法。
Objective To establish a fitting model of hospital days based on BP neural network and to predict the number of hospital days and the sensitivity of influencing factors based on the established neural network model.Using the modeling results of this study for BP neural network Modeling methodology to provide some reference, and to help health management decision makers to make the right decisions and analysis. Methods HIS data was extracted using SQL and modeled and predicted in Clementine 11.1. The prediction results were tested by SPSS 16.0. Results The fitting degree and prediction accuracy of BP neural network were 96.678% and 86.67%, respectively. The number of days of preoperative hospitalization had the greatest impact on the days of hospitalization for radiofrequency catheter ablation. Conclusion Compared with other traditional statistical methods, BP neural network is a suitable modeling method for data characteristics of hospital days.