Grant: Discovery Grants Program - Individual
University of Waterloo
$14,000.00
Latent factor longitudinal social networks and bayesian melding for modelling biological network structures, and longitudinal/spatial bent cables for modelling natural phenomena
FY 2008-09
Summary
Many natural systems and phenomena are highly complex. To gain an adequate understanding of them, scientists develop quantitative techniques to address different aspects of the underlying structures. Observable quantities used in such techniques contain uncertainty, often substantial due to the schemes in which biological data are collected. If such uncertainty is not addressed properly, our understanding of nature would be limited. Statistical modelling can account for such uncertainty. For this proposal, I develop such models, focusing on two major areas: biological network structures, and so-called change-point phenomena.
Biological networks such as food webs have been analyzed quantitatively by (1) descriptive indices defined on the interaction activity levels among species, and (2) so-called computer models that are based on theory in biophysics and/or biochemistry. Other types of biological networks include protein networks in genetical research. Many existing techniques for studying biological network structures do not account for the large amount of uncertainty in the data. My goal is to develop a unifying approach that incorporates bio-physical/chemical theory and statistical modelling principles for complex time-varying network structures. The proposed methodology can address many concerns associated with traditional techniques in biological network research, thus contributing to the advancement of research in this area, as well as the applicability of statistical science to the general scientific community.
Change-point phenomena occur in many natural contexts, e.g. how our climate responds to a major man-made disaster. Traditional techniques assume the change in response is sudden; in my Ph.D. research, I developed a methodology based on the statistical model which I coined the "bent cable," to assess the abruptness of such changes, and found that traditional assumptions are often unfounded, especially in biological contexts where uncertainty is substantial. Since then, I have and will continue to extend the methodology to account for more general data types, including temporally and spatially dependent data.
Program
Discovery Grants Program - Individual
Recipient
Location
Ontario, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
261806-2008
Name nom
Chiu, Grace
Department département
Statistics and Actuarial Science
Organization id
33
Institution établissement
University of Waterloo
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2008
Competition year année de concours
2008
Award amount
14000
Program id
RGPIN
Program nane 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
400
Area of application group en
Environment
Area of application group fr
Environnement
Area of application en
Environment
Area of application fr
Environnement
Research subject code
3000
Research subject group en
Statistics and probability
Research subject group fr
Statistique et probabilité
Research subject en
Statistics and probability
Research subject fr
Statistique et probabilité
Partie
1
Nb partie
2
Application title
Latent factor longitudinal social networks and bayesian melding for modelling biological network structures, and longitudinal/spatial bent cables for modelling natural phenomena
Keyword
NA
Application summary
Many natural systems and phenomena are highly complex. To gain an adequate understanding of them, scientists develop quantitative techniques to address different aspects of the underlying structures. Observable quantities used in such techniques contain uncertainty, often substantial due to the schemes in which biological data are collected. If such uncertainty is not addressed properly, our understanding of nature would be limited. Statistical modelling can account for such uncertainty. For this proposal, I develop such models, focusing on two major areas: biological network structures, and so-called change-point phenomena.
Biological networks such as food webs have been analyzed quantitatively by (1) descriptive indices defined on the interaction activity levels among species, and (2) so-called computer models that are based on theory in biophysics and/or biochemistry. Other types of biological networks include protein networks in genetical research. Many existing techniques for studying biological network structures do not account for the large amount of uncertainty in the data. My goal is to develop a unifying approach that incorporates bio-physical/chemical theory and statistical modelling principles for complex time-varying network structures. The proposed methodology can address many concerns associated with traditional techniques in biological network research, thus contributing to the advancement of research in this area, as well as the applicability of statistical science to the general scientific community.
Change-point phenomena occur in many natural contexts, e.g. how our climate responds to a major man-made disaster. Traditional techniques assume the change in response is sudden; in my Ph.D. research, I developed a methodology based on the statistical model which I coined the "bent cable," to assess the abruptness of such changes, and found that traditional assumptions are often unfounded, especially in biological contexts where uncertainty is substantial. Since then, I have and will continue to extend the methodology to account for more general data types, including temporally and spatially dependent data.