基于LSTM神经网络的喀什地区流腮预测模型  

Kashi district mumps prediction model based on LSTM neural network

作　　者：张志豪 周嘉琪 马国祥 曾婷 ZHANG Zhihao;ZHOU Jiaqi;MA Guoxiang;ZENG Ting(College of Medical Engineering Technology,Xinjiang Medical University,Urumqi 830046,China)

机构地区：[1]新疆医科大学医学工程技术学院,新疆乌鲁木齐830046

出　　处：《现代电子技术》2022年第19期127-132,共6页Modern Electronics Technique

基　　金：新疆维吾尔自治区自然科学基金项目:喀什地区流行性腮腺炎传播机制及疾病定量防控策略研究(2020D01A10)。

摘　　要：流行性腮腺炎(流腮)是一种好发于儿童的急性呼吸道传染病,流行病预测研究工作有助于为有关部门提供科学的辅助决策。以新疆喀什地区2005—2020年流腮病例为研究对象,完成了流行性病学特征分析,分析结果显示,11月为疾病高发时间段,月平均病例数达135例;喀什地区疏附县为病例高发地区,年平均病例达到219例。建立了基于长短期记忆(LSTM)神经网络的喀什地区的流腮预测模型,经过与梯度提升回归树(GBRT)模型对比,实验结果显示LSTM模型的预测精度更高,RMSE误差最小,数值为16.3。基于LSTM模型的流腮预测模型具有良好的预测能力,在实际应用中能够为有关部门提供一定的辅助决策支持。Mumps is a kind of acute respiratory infectious disease which often occurs among children. The study of epidemic prediction is helpful to provide scientific assistant decision-making for relevant departments. The mumps cases in Kashi district in Xinjiang from 2005 to 2020 are taken as the research object to analyze the epidemiological characteristics. Results of the analysis show that the November is the high incidence period of the disease,and the monthly average number of cases in the November is up to 135,and that the Shufu County in Kashi district is a high incidence area,with an annual average number of219 cases. A Kashi district mumps prediction model based on long and short-term memory(LSTM)neural network is established and compared with the gradient boosting regression tree(GBRT) model. The experimental results show that the prediction accuracy of the LSTM model is higher,and its RMSE is the smaller,whose value is 16.3. The mumps prediction model based on the LSTM model has good prediction ability,and can provide some auxiliary decision support for relevant departments in practical application.

关 键 词：流腮预测模型 时间序列 LSTM神经网络 GBRT模型 流行病学特征分析 时间特征分析 空间特征分析 

分 类 号：TN919-34[电子电信—通信与信息系统] R183[电子电信—信息与通信工程]