{"title":"利用神经网络预测具有相关不确定性的传染病流行率","authors":"Michael Morris","doi":"arxiv-2409.01154","DOIUrl":null,"url":null,"abstract":"Infectious diseases pose significant human and economic burdens. Accurately\nforecasting disease incidence can enable public health agencies to respond\neffectively to existing or emerging diseases. Despite progress in the field,\ndeveloping accurate forecasting models remains a significant challenge. This\nthesis proposes two methodological frameworks using neural networks (NNs) with\nassociated uncertainty estimates - a critical component limiting the\napplication of NNs to epidemic forecasting thus far. We develop our frameworks\nby forecasting influenza-like illness (ILI) in the United States. Our first\nproposed method uses Web search activity data in conjunction with historical\nILI rates as observations for training NN architectures. Our models incorporate\nBayesian layers to produce uncertainty intervals, positioning themselves as\nlegitimate alternatives to more conventional approaches. The best performing\narchitecture: iterative recurrent neural network (IRNN), reduces mean absolute\nerror by 10.3% and improves Skill by 17.1% on average in forecasting tasks\nacross four flu seasons compared to the state-of-the-art. We build on this\nmethod by introducing IRNNs, an architecture which changes the sampling\nprocedure in the IRNN to improve the uncertainty estimation. Our second\nframework uses neural ordinary differential equations to bridge the gap between\nmechanistic compartmental models and NNs; benefiting from the physical\nconstraints that compartmental models provide. We evaluate eight neural ODE\nmodels utilising a mixture of ILI rates and Web search activity data to provide\nforecasts. These are compared with the IRNN and IRNN0 - the IRNN using only ILI\nrates. Models trained without Web search activity data outperform the IRNN0 by\n16% in terms of Skill. Future work should focus on more effectively using\nneural ODEs with Web search data to compete with the best performing IRNN.","PeriodicalId":501044,"journal":{"name":"arXiv - QuanBio - Populations and Evolution","volume":"16 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Forecasting infectious disease prevalence with associated uncertainty using neural networks\",\"authors\":\"Michael Morris\",\"doi\":\"arxiv-2409.01154\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Infectious diseases pose significant human and economic burdens. Accurately\\nforecasting disease incidence can enable public health agencies to respond\\neffectively to existing or emerging diseases. Despite progress in the field,\\ndeveloping accurate forecasting models remains a significant challenge. This\\nthesis proposes two methodological frameworks using neural networks (NNs) with\\nassociated uncertainty estimates - a critical component limiting the\\napplication of NNs to epidemic forecasting thus far. We develop our frameworks\\nby forecasting influenza-like illness (ILI) in the United States. Our first\\nproposed method uses Web search activity data in conjunction with historical\\nILI rates as observations for training NN architectures. Our models incorporate\\nBayesian layers to produce uncertainty intervals, positioning themselves as\\nlegitimate alternatives to more conventional approaches. The best performing\\narchitecture: iterative recurrent neural network (IRNN), reduces mean absolute\\nerror by 10.3% and improves Skill by 17.1% on average in forecasting tasks\\nacross four flu seasons compared to the state-of-the-art. We build on this\\nmethod by introducing IRNNs, an architecture which changes the sampling\\nprocedure in the IRNN to improve the uncertainty estimation. Our second\\nframework uses neural ordinary differential equations to bridge the gap between\\nmechanistic compartmental models and NNs; benefiting from the physical\\nconstraints that compartmental models provide. We evaluate eight neural ODE\\nmodels utilising a mixture of ILI rates and Web search activity data to provide\\nforecasts. These are compared with the IRNN and IRNN0 - the IRNN using only ILI\\nrates. Models trained without Web search activity data outperform the IRNN0 by\\n16% in terms of Skill. Future work should focus on more effectively using\\nneural ODEs with Web search data to compete with the best performing IRNN.\",\"PeriodicalId\":501044,\"journal\":{\"name\":\"arXiv - QuanBio - Populations and Evolution\",\"volume\":\"16 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-02\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - QuanBio - Populations and Evolution\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.01154\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - QuanBio - Populations and Evolution","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.01154","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
摘要
传染病给人类和经济造成了巨大负担。准确预测疾病的发病率可以使公共卫生机构有效应对现有的或新出现的疾病。尽管该领域取得了进展,但开发准确的预测模型仍是一项重大挑战。本论文提出了两个使用神经网络(NN)的方法框架以及相关的不确定性估计,这是迄今为止限制神经网络应用于流行病预测的一个关键因素。我们通过预测美国的流感样疾病(ILI)来开发我们的框架。我们提出的第一种方法将网络搜索活动数据与历史 ILI 发病率结合起来,作为训练 NN 架构的观测数据。我们的模型结合贝叶斯层来产生不确定性区间,将自己定位为传统方法的合法替代品。在四个流感季节的预测任务中,表现最好的架构:迭代递归神经网络(IRNN)与最先进的架构相比,平均绝对误差减少了 10.3%,技能提高了 17.1%。我们在这一方法的基础上引入了 IRNN,这一架构改变了 IRNN 中的采样过程,从而改进了不确定性估计。我们的第二个框架使用神经常微分方程来弥合机理分区模型和神经网络之间的差距,并从分区模型提供的物理约束中获益。我们利用 ILI 率和网络搜索活动数据的混合物来提供预测,并对八个神经 ODE 模型进行了评估。这些模型与 IRNN 和 IRNN0(IRNN 仅使用 ILI 率)进行了比较。不使用网络搜索活动数据训练的模型在技能方面比 IRNN0 高出 16%。未来的工作重点应该是更有效地利用网络搜索数据来使用神经 ODE,从而与表现最好的 IRNN 竞争。
Forecasting infectious disease prevalence with associated uncertainty using neural networks
Infectious diseases pose significant human and economic burdens. Accurately
forecasting disease incidence can enable public health agencies to respond
effectively to existing or emerging diseases. Despite progress in the field,
developing accurate forecasting models remains a significant challenge. This
thesis proposes two methodological frameworks using neural networks (NNs) with
associated uncertainty estimates - a critical component limiting the
application of NNs to epidemic forecasting thus far. We develop our frameworks
by forecasting influenza-like illness (ILI) in the United States. Our first
proposed method uses Web search activity data in conjunction with historical
ILI rates as observations for training NN architectures. Our models incorporate
Bayesian layers to produce uncertainty intervals, positioning themselves as
legitimate alternatives to more conventional approaches. The best performing
architecture: iterative recurrent neural network (IRNN), reduces mean absolute
error by 10.3% and improves Skill by 17.1% on average in forecasting tasks
across four flu seasons compared to the state-of-the-art. We build on this
method by introducing IRNNs, an architecture which changes the sampling
procedure in the IRNN to improve the uncertainty estimation. Our second
framework uses neural ordinary differential equations to bridge the gap between
mechanistic compartmental models and NNs; benefiting from the physical
constraints that compartmental models provide. We evaluate eight neural ODE
models utilising a mixture of ILI rates and Web search activity data to provide
forecasts. These are compared with the IRNN and IRNN0 - the IRNN using only ILI
rates. Models trained without Web search activity data outperform the IRNN0 by
16% in terms of Skill. Future work should focus on more effectively using
neural ODEs with Web search data to compete with the best performing IRNN.