Canada Spends Logo

Explorez notre analyse du budget 2026

Grant: Discovery Grants Program - Individual

University of Western Ontario

$19,000.00

Inference in the presence of constraints

EF 2010-11

Résumé
One of the main goals of this project is to improve curve and surface estimation when data are limited but when input from scientific experts is available. A number of smoothing techniques exist for flexibly modeling data, and these techniques have good properties. It is possible to apply these techniques to forms of data which have been adjusted according to previously available science. Depending on the accuracy of the science, the resulting estimates should be better than those obtained from the unadjusted data. The technique of modifying the data, taking into account qualitative information, is called data sharpening. By employing data sharpening, scientists should be able to obtain improved estimates of quantities they are interested in (usually in the form of curves and surfaces), and they should also be able to obtain better predictions as a result. Software for data sharpening will be developed in order to handle a wide variety of modeling problems. The project will also be concerned with incomplete forms of data, such as where data are not measured precisely but are instead reported in the form of an interval which contains the true measurement. Again, we are interested in estimating curves and surfaces. The data sharpening method will be applied here when there is additional qualitative information about the curves or surfaces, in addition to iterative methods that are under development. Methods of resampling data are also being studied; these methods (called bootstrap methods) are used to estimate variability in parameter estimates in situations where closed form mathematical formulae are not available. Some of the statistical methods being developed in this proposal will find direct application in forest fire management and in climate modeling. In particular, smoothing methods and bootstrap methods will be applied to input data for a deterministic forest fire spread model, leading to probability maps of fire occurrence. Smoothing-based clustering algorithms will be applied to storm systems in order to aid in modeling and prediction.
Programme
Discovery Grants Program - Individual
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data

Tel que publié

Application id
138127-2009
Name nom
Braun, Willard
Department département
Statistical and Actuarial Sciences
Organization id
36
Institution établissement
University of Western Ontario
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2010
Competition year année de concours
2009
Award amount
19000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
Discovery grants programs
Group fr
Programmes de subventions à la découverte
Committee code
14
Committee name en
Statistical Sciences
Committee name fr
Statistique
Area of application code
201
Area of application group en
Natural resources (economic aspects)
Area of application group fr
Ressources naturelles (aspects économiques)
Area of application en
Forestry (silviculture, forest management)
Area of application fr
Foresterie (sylviculture, aménagement forestier)
Research subject code
3010
Research subject group en
Statistics and probability
Research subject group fr
Statistique et probabilité
Research subject en
Nonparametric inference
Research subject fr
Inférence non paramétrique
Partie
2
Nb partie
5
Application title
Inference in the presence of constraints
Keyword
NSERC
Application summary
One of the main goals of this project is to improve curve and surface estimation when data are limited but when input from scientific experts is available. A number of smoothing techniques exist for flexibly modeling data, and these techniques have good properties. It is possible to apply these techniques to forms of data which have been adjusted according to previously available science. Depending on the accuracy of the science, the resulting estimates should be better than those obtained from the unadjusted data. The technique of modifying the data, taking into account qualitative information, is called data sharpening. By employing data sharpening, scientists should be able to obtain improved estimates of quantities they are interested in (usually in the form of curves and surfaces), and they should also be able to obtain better predictions as a result. Software for data sharpening will be developed in order to handle a wide variety of modeling problems. The project will also be concerned with incomplete forms of data, such as where data are not measured precisely but are instead reported in the form of an interval which contains the true measurement. Again, we are interested in estimating curves and surfaces. The data sharpening method will be applied here when there is additional qualitative information about the curves or surfaces, in addition to iterative methods that are under development. Methods of resampling data are also being studied; these methods (called bootstrap methods) are used to estimate variability in parameter estimates in situations where closed form mathematical formulae are not available. Some of the statistical methods being developed in this proposal will find direct application in forest fire management and in climate modeling. In particular, smoothing methods and bootstrap methods will be applied to input data for a deterministic forest fire spread model, leading to probability maps of fire occurrence. Smoothing-based clustering algorithms will be applied to storm systems in order to aid in modeling and prediction.