Grant: Alliance Grants
Université de Montréal
$45,000.00
High-dimensional hierarchical data with complex dependence structures
EF 2024-25
Résumé
The principal aim of this research project is to predict an outcome based on covariables when data are grouped at multiple levels or possesses a predefined hierarchical structure. Such data naturally arises in various research fields. For instance, the impact of a heatwave on an individual's health varies, influenced by factors such as their place of residence, financial resources, health-related variables, etc. Indeed, for the same health conditions, residents of underprivileged neighborhoods are more likely to be affected by heatwaves than those in affluent neighborhoods. Since weather conditions are local and non-uniform, geographical location is also significant. Consequently, dividing the target population into multiple groups is appropriate based on various factors. The most frequently employed statistical tool to analyze this type of data is a multilevel linear regression, which provides elegant means to estimate a factor's effect on the outcome variable. In the previous example, this approach makes it possible to measure the impact of residing in a specific neighborhood on residents' health. Nevertheless, this approach, owing to its simplicity, fails to accommodate the intricate interplay of variables. Over the past decade, regression methods based on dependence functions, called copulas, have emerged with highly promising results. This project aims to utilize copulas for modeling the dependence between an outcome variable, covariates, and the dependency induced by the factors. Copulas will also be employed to introduce dependence into the hierarchical structure. The primary goal of this research project is to develop inference methods for copula regression in high-dimensional hierarchical data. Our main target applications are in the fields of health, environmental studies, and climate change. Furthermore, we will create libraries using the open-source statistical software R to enhance the accessibility or our developed methods.
Programme
Alliance Grants
Bénéficiaire
Emplacement
Quebec, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
586845-2023
Name nom
Nasri, Bouchra BN
Department département
Médecine sociale et préventive
Organization id
63
Institution établissement
Université de Montréal
Province en
Québec
Province fr
Québec
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2023
Award amount
45000
Program id
ALLRP
Program name en
Alliance Grants
Program name fr
Subventions Alliance
Group en
RESEARCH PARTNERSHIPS
Group fr
PARTENARIATS DE RECHERCHE
Committee code
96
Committee name en
RPP Internal Decision Cttee
Committee name fr
Cté de décision interne - PPR
Area of application code
401
Area of application group en
Environment
Area of application group fr
Environnement
Area of application en
Climate and atmosphere
Area of application fr
Climat et atmosphère
Research subject code
3011
Research subject group en
Statistics and probability
Research subject group fr
Statistique et probabilité
Research subject en
Multivariate analysis
Research subject fr
Analyse multivariée
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
High-dimensional hierarchical data with complex dependence structures
Keywords
copulas; dependence modelling; hierarchical data; high-dimensional data; latent factors; mixed effect models
Application summary
The principal aim of this research project is to predict an outcome based on covariables when data are grouped at multiple levels or possesses a predefined hierarchical structure. Such data naturally arises in various research fields. For instance, the impact of a heatwave on an individual's health varies, influenced by factors such as their place of residence, financial resources, health-related variables, etc. Indeed, for the same health conditions, residents of underprivileged neighborhoods are more likely to be affected by heatwaves than those in affluent neighborhoods. Since weather conditions are local and non-uniform, geographical location is also significant. Consequently, dividing the target population into multiple groups is appropriate based on various factors. The most frequently employed statistical tool to analyze this type of data is a multilevel linear regression, which provides elegant means to estimate a factor's effect on the outcome variable. In the previous example, this approach makes it possible to measure the impact of residing in a specific neighborhood on residents' health. Nevertheless, this approach, owing to its simplicity, fails to accommodate the intricate interplay of variables. Over the past decade, regression methods based on dependence functions, called copulas, have emerged with highly promising results. This project aims to utilize copulas for modeling the dependence between an outcome variable, covariates, and the dependency induced by the factors. Copulas will also be employed to introduce dependence into the hierarchical structure. The primary goal of this research project is to develop inference methods for copula regression in high-dimensional hierarchical data. Our main target applications are in the fields of health, environmental studies, and climate change. Furthermore, we will create libraries using the open-source statistical software R to enhance the accessibility or our developed methods.