서브메뉴
검색
Bayesian Optimization With Gaussian Process Emulation Applied to Calibration of Environmental and Within-Host Parameters in an Agent-Based Model of Plasmodium falciparum Malaria Transmission
Bayesian Optimization With Gaussian Process Emulation Applied to Calibration of Environmental and Within-Host Parameters in an Agent-Based Model of Plasmodium falciparum Malaria Transmission
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202103522
- ISBN
- 9798315797388
- DDC
- 614.4
- 서명/저자
- Bayesian Optimization With Gaussian Process Emulation Applied to Calibration of Environmental and Within-Host Parameters in an Agent-Based Model of Plasmodium falciparum Malaria Transmission
- 발행사항
- [Sl] : Northwestern University, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 103 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 86-12, Section: B.
- 주기사항
- Advisor: Gerardin, Jaline.
- 학위논문주기
- Thesis (Ph.D.)--Northwestern University, 2025.
- 초록/해제
- 요약Malaria continues to be a major global health concern and is responsible for more than 500,000 deaths annually, with more than 90% in resource-limited areas of sub-Saharan Africa. Individual-based models have become key parts of the effort to eradicate malaria, informing decisions about when, where, and how to effectively intervene against disease. These models simulate processes within and interaction between humans and mosquitos such that population-level phenomena emerge mechanistically but require calibration to site-specific epidemiological data following changes to model structure. For more detailed, complex models, the large number of uncertain model parameters makes calibration to epidemiological data challenging. We developed a user-friendly software pipeline for applying Bayesian optimization with Gaussian process emulation and trust-region based Thompson sampling to calibration of a complex malaria transmission model. We demonstrated the functionality of the pipeline in two key calibration applications. The pipeline successfully captured local malaria transmission intensity and seasonality from primary data in a mechanistic model of climate-driven mosquito development, generating a baseline model of transmission at a well-studied malaria trial site. Following updates to introduce heterogeneity and decay to modeled immunity, the pipeline rescued model goodness-of-fit to data describing uncomplicated and severe disease incidence, prevalence, parasite densities, and infectiousness across eight diverse malaria-endemic sites. The resulting recalibrated within-host model also had improved fit to out-of-sample datasets describing infection duration and multiplicity.
- 일반주제명
- Epidemiology
- 일반주제명
- Applied mathematics
- 일반주제명
- Parasitology
- 키워드
- Calibration
- 키워드
- Malaria
- 키워드
- Modeling
- 키워드
- Optimization
- 키워드
- Seasonality
- 기타저자
- Northwestern University Health Sciences Integrated PhD Program
- 기본자료저록
- Dissertations Abstracts International. 86-12B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2025 us c eng d■001000017357510
■00520260202103522
■006m o d
■007cr#unu||||||||
■020 ▼a9798315797388
■035 ▼a(MiAaPQ)AAI32038837
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a614.4
■1001 ▼aHolden, Tobias McKenzie.▼0(orcid)0000-0002-0793-9585
■24510▼aBayesian Optimization With Gaussian Process Emulation Applied to Calibration of Environmental and Within-Host Parameters in an Agent-Based Model of Plasmodium falciparum Malaria Transmission
■260 ▼a[Sl]▼bNorthwestern University▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a103 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 86-12, Section: B.
■500 ▼aAdvisor: Gerardin, Jaline.
■5021 ▼aThesis (Ph.D.)--Northwestern University, 2025.
■520 ▼aMalaria continues to be a major global health concern and is responsible for more than 500,000 deaths annually, with more than 90% in resource-limited areas of sub-Saharan Africa. Individual-based models have become key parts of the effort to eradicate malaria, informing decisions about when, where, and how to effectively intervene against disease. These models simulate processes within and interaction between humans and mosquitos such that population-level phenomena emerge mechanistically but require calibration to site-specific epidemiological data following changes to model structure. For more detailed, complex models, the large number of uncertain model parameters makes calibration to epidemiological data challenging. We developed a user-friendly software pipeline for applying Bayesian optimization with Gaussian process emulation and trust-region based Thompson sampling to calibration of a complex malaria transmission model. We demonstrated the functionality of the pipeline in two key calibration applications. The pipeline successfully captured local malaria transmission intensity and seasonality from primary data in a mechanistic model of climate-driven mosquito development, generating a baseline model of transmission at a well-studied malaria trial site. Following updates to introduce heterogeneity and decay to modeled immunity, the pipeline rescued model goodness-of-fit to data describing uncomplicated and severe disease incidence, prevalence, parasite densities, and infectiousness across eight diverse malaria-endemic sites. The resulting recalibrated within-host model also had improved fit to out-of-sample datasets describing infection duration and multiplicity.
■590 ▼aSchool code: 0163.
■650 4▼aEpidemiology
■650 4▼aApplied mathematics
■650 4▼aParasitology
■653 ▼aCalibration
■653 ▼aMalaria
■653 ▼aModeling
■653 ▼aOptimization
■653 ▼aSeasonality
■653 ▼aWithin-host model
■690 ▼a0766
■690 ▼a0364
■690 ▼a0718
■71020▼aNorthwestern University▼bHealth Sciences Integrated PhD Program.
■7730 ▼tDissertations Abstracts International▼g86-12B.
■790 ▼a0163
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17357510▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
