Grant: Discovery Grants Program - Individual
The University of British Columbia
$33,000.00
Robust estimation methods for complex data models
EF 2023-24
Résumé
The overarching goal of my research program is to develop and study statistical methodology that is relevant and feasible for applications with large and complex data, and that remains reliable even when a portion of the data may have been affected by atypical errors (recording mistakes, etc.) Since data collection is now often relatively easy and inexpensive, many data sets are collected in an "opportunistic" manner, in other words, they are obtained without traditional experimental design considerations, or perhaps even without a specific target population in mind. Consequently, model adequacy and the potential bias resulting from model misspecification become important concerns that need to be addressed in order for assess the validity of our conclusions. The problem to which I refer here is not the usual model selection issue of finding which of a collection of potential models is most appropriate for the data, but rather assessing whether a model fits most of the observations very well at the expense of possibly not accommodating a small proportion of points. These "atypical points" need not be extreme observations, but perhaps points coming from a different population which were included due to having cast a very "wide" net to collect our data. Statistical methods that can deal with this issue efficiently will play an increasingly important role in the near future.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
British Columbia, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
RGPIN-2023-04489
Name nom
Salibian Barrera, Matias
Department département
Department of Statistics
Organization id
2
Institution établissement
The University of British Columbia
Province en
British Columbia
Province fr
Colombie-Britannique
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2023
Competition year année de concours
2023
Award amount
33000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1508
Committee name en
Mathematics and Statistics
Committee name fr
Mathématiques et statistique
Area of application code
1100
Area of application group en
Health, education and social services
Area of application group fr
Santé publique, éducation et services sociaux
Area of application en
Health, education and social services
Area of application fr
Santé publique, éducation et services sociaux
Research subject code
3001
Research subject group en
Statistics and probability
Research subject group fr
Statistical theory
Research subject en
Statistique et probabilité
Research subject fr
Théorie de statistique
Application title
Robust estimation methods for complex data models
Keyword
High-dimensional models; Non-parametrics; Robust Statistics
Application summary
The overarching goal of my research program is to develop and study statistical methodology that is relevant and feasible for applications with large and complex data, and that remains reliable even when a portion of the data may have been affected by atypical errors (recording mistakes, etc.) Since data collection is now often relatively easy and inexpensive, many data sets are collected in an "opportunistic" manner, in other words, they are obtained without traditional experimental design considerations, or perhaps even without a specific target population in mind. Consequently, model adequacy and the potential bias resulting from model misspecification become important concerns that need to be addressed in order for assess the validity of our conclusions. The problem to which I refer here is not the usual model selection issue of finding which of a collection of potential models is most appropriate for the data, but rather assessing whether a model fits most of the observations very well at the expense of possibly not accommodating a small proportion of points. These "atypical points" need not be extreme observations, but perhaps points coming from a different population which were included due to having cast a very "wide" net to collect our data. Statistical methods that can deal with this issue efficiently will play an increasingly important role in the near future.