数据驱动的心血管疾病门诊量多步组合预测研究  

Research on Data-driven Multi-step Combined Forecast of Cardiovascular Disease Outpatient Volume

作　　者：顾福来[1] 白朝阳[1,2] 郭林霞 刘晓冰[1,2] 孙永亮 GU Fulai;BAI Zhaoyang;GUO Linxia;LIU Xiaobing;SUN Yongliang(School of Economics and Management,Dalian University of Technology,Dalian 116024,China;National and Local Joint Engineering Laboratory for Manufacturing Management Information Technology,Dalian 116024,China)

机构地区：[1]大连理工大学经济管理学院,辽宁大连116024 [2]大连理工大学制造管理信息化技术国家地方联合工程实验室,辽宁大连116024

出　　处：《信息系统学报》2022年第1期19-31,共13页China Journal of Information Systems

摘　　要：精准的心血管门诊量预测是实现医院医生需求计算、医疗设备分配和管理的重要基础。本文基于心血管门诊量时间序列数据,采用迭代策略进行多步长时间序列预测,为降低多步长预测带来的误差累积问题和数据非平稳、非线性的特征,提出改进集合经验模态分解方法,并结合反向传播神经网络(backpropagationneuralnetwork,BPNN)建立组合预测模型,达到更优的预测结果。最后以某医院心血管疾病月门诊量进行预测对比分析,实验结果表明,该组合预测模型对心血管门诊量多步长预测准确率较高,证实了模型的有效性。Precise cardiovascular outpatient volume forecasting is an important basis for realizing outpatient management such as doctor demand calculations and medical equipment management and allocation.In this paper,based on the time series data of cardiovascular outpatient volume,multi-step time series prediction is carried out by iterative strategy.In order to reduce the error accumulation caused by multi-step prediction and the non-stationary and non-linear characteristics of data,an improved ensemble empirical mode decomposition method is proposed,and a combined prediction model is established with back propagation neural network(BPNN)to achieve better prediction results.Finally,the monthly outpatient volume of cardiovascular diseases is used for prediction and comparative analysis.The experimental results show that the combined prediction model has high accuracy in predicting the cardiovascular outpatient volume with multiple steps,which proves the effectiveness of the model.

关 键 词：心血管门诊量 多步预测 改进集合经验模态分解 BPNN 

分 类 号：G311[文化科学]