{"title":"传染病建模中的基于代理的模型与区室模型的整合:一种新颖的混合方法","authors":"Inan Bostanci, Tim Conrad","doi":"arxiv-2407.20993","DOIUrl":null,"url":null,"abstract":"This study investigates the spatial integration of agent-based models (ABMs)\nand compartmental models in infectious disease modeling, presenting a novel\nhybrid approach and studying its implications. ABMs, characterized by\nindividual agent interactions and decision-making, offer detailed insights but\nare computationally intensive for large populations. Compartmental models,\nbased on differential equations, provide population-level dynamics but lack\ngranular detail. Our hybrid model aims to balance the granularity of ABMs with\nthe computational efficiency of compartmental models, offering a more nuanced\nunderstanding of disease spread in diverse scenarios, including large\npopulations. We developed a custom ABM and a compartmental model, analyzing\ntheir infectious disease dynamics separately before integrating them into a\nhybrid model. This integration involved spatial coupling of discrete and\ncontinuous populations and evaluating the consistency of disease dynamics at\nthe macro scale. Our key objectives were to assess the effect of model\nhybridization on resulting infection dynamics, and to quantify computational\ncost savings of the hybrid approach over the ABM. We show that the hybrid\napproach can significantly reduce computational costs, but is sensitive to\nbetween-model differences, highlighting that model equivalence is a crucial\ncomponent of hybrid modeling approaches. The code is available at\nhttp://github.com/iebos/hybrid_model1.","PeriodicalId":501044,"journal":{"name":"arXiv - QuanBio - Populations and Evolution","volume":"31 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Integrating Agent-Based and Compartmental Models for Infectious Disease Modeling: A Novel Hybrid Approach\",\"authors\":\"Inan Bostanci, Tim Conrad\",\"doi\":\"arxiv-2407.20993\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"This study investigates the spatial integration of agent-based models (ABMs)\\nand compartmental models in infectious disease modeling, presenting a novel\\nhybrid approach and studying its implications. ABMs, characterized by\\nindividual agent interactions and decision-making, offer detailed insights but\\nare computationally intensive for large populations. Compartmental models,\\nbased on differential equations, provide population-level dynamics but lack\\ngranular detail. Our hybrid model aims to balance the granularity of ABMs with\\nthe computational efficiency of compartmental models, offering a more nuanced\\nunderstanding of disease spread in diverse scenarios, including large\\npopulations. We developed a custom ABM and a compartmental model, analyzing\\ntheir infectious disease dynamics separately before integrating them into a\\nhybrid model. This integration involved spatial coupling of discrete and\\ncontinuous populations and evaluating the consistency of disease dynamics at\\nthe macro scale. Our key objectives were to assess the effect of model\\nhybridization on resulting infection dynamics, and to quantify computational\\ncost savings of the hybrid approach over the ABM. We show that the hybrid\\napproach can significantly reduce computational costs, but is sensitive to\\nbetween-model differences, highlighting that model equivalence is a crucial\\ncomponent of hybrid modeling approaches. The code is available at\\nhttp://github.com/iebos/hybrid_model1.\",\"PeriodicalId\":501044,\"journal\":{\"name\":\"arXiv - QuanBio - Populations and Evolution\",\"volume\":\"31 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-07-30\",\"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-2407.20993\",\"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-2407.20993","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Integrating Agent-Based and Compartmental Models for Infectious Disease Modeling: A Novel Hybrid Approach
This study investigates the spatial integration of agent-based models (ABMs)
and compartmental models in infectious disease modeling, presenting a novel
hybrid approach and studying its implications. ABMs, characterized by
individual agent interactions and decision-making, offer detailed insights but
are computationally intensive for large populations. Compartmental models,
based on differential equations, provide population-level dynamics but lack
granular detail. Our hybrid model aims to balance the granularity of ABMs with
the computational efficiency of compartmental models, offering a more nuanced
understanding of disease spread in diverse scenarios, including large
populations. We developed a custom ABM and a compartmental model, analyzing
their infectious disease dynamics separately before integrating them into a
hybrid model. This integration involved spatial coupling of discrete and
continuous populations and evaluating the consistency of disease dynamics at
the macro scale. Our key objectives were to assess the effect of model
hybridization on resulting infection dynamics, and to quantify computational
cost savings of the hybrid approach over the ABM. We show that the hybrid
approach can significantly reduce computational costs, but is sensitive to
between-model differences, highlighting that model equivalence is a crucial
component of hybrid modeling approaches. The code is available at
http://github.com/iebos/hybrid_model1.