Canada Spends Logo

Explore our Budget 2026 analysis

Grant: Discovery Grants Program - Individual

Queen's University

$38,000.00

Novel statistical models and learning methods for complex event history data

FY 2024-25

Summary
Event history data include censored survival, longitudinal repeated measurements, and recurrent event data. They are often found in many disciplines. For example, censored failure time data from stress tests for reliability in engineering, censored time to employment in econometrics, censored time to bank bankruptcy in mathematical finance, longitudinal biomarkers and censored survival times in humans and species in medicine and biology, and so on. Developing novel models and estimation methods for such data, including cases where some subjects may be immune to the event of interest (called cured subjects), is the focus of my research program. The overarching goal of this study is to develop novel approaches for the data so that fewer parametric assumptions will be required, more covariates will be allowed, more complicated covariate effects including causal effects will be identified and estimated, and cured subjects will be better handled in modeling the data. The fast development of machine learning methods in the last decade provides an opportunity to exploit the methods in estimating statistical models for event history data and develop new learning methods for the data that are efficient for big data and enjoy sound statistical properties. Short-term objectives in the next 5 years include but not limited to - examining the limitations of the existing nonparametric estimation method and developing novel machine learning methods for the mixture cure models that can handle high-dimensional covariates and more complex covariate effects for right-censored and interval-censored survival data. - proposing a framework and conditions for causal inference for cure models and developing estimation methods to estimate causal treatment effects. - developing novel machine learning methods to explore treatment heterogeneity for survival data with a cured fraction and longitudinal data. - addressing issues in the existing estimation methods for the quantile regression-based cure models for survival data with a cured fraction and proposing novel estimation methods for flexible quantile cure models with high-dimensional data. The proposed research program will provide a multidisciplinary training environment for trainees to develop novel statistical models that are highly relevant in analyzing event history data. We will recruit and train MSc students, PhD students, and a post-doctoral fellow in this program. The results of this research program are expected to be disseminated by publications in high-quality statistical journals, readily usable software packages, and presentations in conferences and seminars. The proposed research will contribute substantially to the advances of statistical methodology research and applications and will be able to provide solutions for complex effects in big data that are actionable and causally effective.
Program
Discovery Grants Program - Individual
Location
Ontario, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data

As published

Application id
RGPIN-2024-04481
Name nom
Peng, Yingwei
Department département
Public Health Sciences/Mathematics and Statistics
Organization id
29
Institution établissement
Queen's University
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2024
Award amount
38000
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
1201
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Life sciences (including biotechnology)
Area of application fr
Sciences de la vie (y compris biotechnologie)
Research subject code
3005
Research subject group en
Statistics and probability
Research subject group fr
Statistique et probabilité
Research subject en
Biostatistics
Research subject fr
Biostatistiques
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
Novel statistical models and learning methods for complex event history data
Keywords
survival models; censored data; longitudinal data; cure models; machine learning methods; causal inference; statistical methods for big data; joint modeling
Application summary
Event history data include censored survival, longitudinal repeated measurements, and recurrent event data. They are often found in many disciplines. For example, censored failure time data from stress tests for reliability in engineering, censored time to employment in econometrics, censored time to bank bankruptcy in mathematical finance, longitudinal biomarkers and censored survival times in humans and species in medicine and biology, and so on. Developing novel models and estimation methods for such data, including cases where some subjects may be immune to the event of interest (called cured subjects), is the focus of my research program. The overarching goal of this study is to develop novel approaches for the data so that fewer parametric assumptions will be required, more covariates will be allowed, more complicated covariate effects including causal effects will be identified and estimated, and cured subjects will be better handled in modeling the data. The fast development of machine learning methods in the last decade provides an opportunity to exploit the methods in estimating statistical models for event history data and develop new learning methods for the data that are efficient for big data and enjoy sound statistical properties. Short-term objectives in the next 5 years include but not limited to - examining the limitations of the existing nonparametric estimation method and developing novel machine learning methods for the mixture cure models that can handle high-dimensional covariates and more complex covariate effects for right-censored and interval-censored survival data. - proposing a framework and conditions for causal inference for cure models and developing estimation methods to estimate causal treatment effects. - developing novel machine learning methods to explore treatment heterogeneity for survival data with a cured fraction and longitudinal data. - addressing issues in the existing estimation methods for the quantile regression-based cure models for survival data with a cured fraction and proposing novel estimation methods for flexible quantile cure models with high-dimensional data. The proposed research program will provide a multidisciplinary training environment for trainees to develop novel statistical models that are highly relevant in analyzing event history data. We will recruit and train MSc students, PhD students, and a post-doctoral fellow in this program. The results of this research program are expected to be disseminated by publications in high-quality statistical journals, readily usable software packages, and presentations in conferences and seminars. The proposed research will contribute substantially to the advances of statistical methodology research and applications and will be able to provide solutions for complex effects in big data that are actionable and causally effective.
Queen's University - $38,000 Grant (2024-25) | Canada Spends