Grant: Discovery Grants Program - Individual
University of Waterloo
$33,000.00
Theory and Applications of Functional Time Series Analysis
EF 2023-24
Résumé
In critical areas of research ranging from environmental studies to cancer, large amounts of data are collected continuously over time on objects that can best be described as curves or surfaces. Such data are termed functional time series (FTS). Examples include dense temperature records fashioned into daily temperature curves or surfaces, which are used to study changes and anomalies in climate, or sequentially observed density functions describing the motion of immune cells interacting with cancer cells that aid in understanding the efficacy of new immunotherapies. The goal of the proposed research is to advance knowledge of functional time series analysis (FTSA), an emerging sub-field of statistics built to provide insights into such data. The research outlined in this proposal expands the knowledge of FTSA in five primary directions. Change Point Analysis: To aid in the modelling of FTS, one often must evaluate whether their random behavior appears to stay stable. When this is not the case, it is of interest to perform change point analysis, in which the series is divided into more homogeneous segments. We aim to develop and expand general change point procedures for FTS. These will importantly make use of global selection criteria tailored to functional data for performing multiple change point analysis, as well as robust approaches to tune these methods. White Noise Tests: A classical problem, crucial in developing models for FTS, is to evaluate whether a time series contains any meaningful "signal", or instead appears to evolve as pure noise. We will extend white noise tests for FTS in several important directions, including the evaluation of the goodness-of-fit of oft-used FTS models. We also develop methods for detecting non-standard signals in FTS not captured by existing approaches. Volatility Models: Models for volatile FTS provide novel ways of incorporating high frequency asset price data to measure financial risk. These models so far rely on key simplifying assumptions that severely limit their flexibility. We will correct these deficiencies by proposing new, more complete, volatility models for FTS, along with accompanying estimation and model identification procedures. Robust Methods: Essentially all existing methods to study FTS are highly susceptible to the influence of outliers or irregularities in the data. We aim to develop companion robust alternatives to existing methods in FTSA, notably to measure autocorrelation and perform dimension reduction tailored to FTS. Applications to Cancer Cell Biology and Immunotherapy: Imaging data from cancer biopsies using modern microscopes capture the motion of immune cells as they interact with cancer cells. Based on these data, we will develop methods in FTSA useful to improve our understanding of interactions between immune cells and cancer cells, and to predict success of emerging therapies.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
RGPIN-2023-03485
Name nom
Rice, Greg
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
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
1205
Area of application group en
Advancement of knowledge
Area of application group fr
Avancement des connaissances
Area of application en
Mathematical sciences
Area of application fr
Sciences mathématiques
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
Theory and Applications of Functional Time Series Analysis
Keyword
Change Point Analysis; Functional Data Analysis; GARCH; Model Selection; Robustness; Volatility Models; White Noise Testing
Application summary
In critical areas of research ranging from environmental studies to cancer, large amounts of data are collected continuously over time on objects that can best be described as curves or surfaces. Such data are termed functional time series (FTS). Examples include dense temperature records fashioned into daily temperature curves or surfaces, which are used to study changes and anomalies in climate, or sequentially observed density functions describing the motion of immune cells interacting with cancer cells that aid in understanding the efficacy of new immunotherapies. The goal of the proposed research is to advance knowledge of functional time series analysis (FTSA), an emerging sub-field of statistics built to provide insights into such data. The research outlined in this proposal expands the knowledge of FTSA in five primary directions. Change Point Analysis: To aid in the modelling of FTS, one often must evaluate whether their random behavior appears to stay stable. When this is not the case, it is of interest to perform change point analysis, in which the series is divided into more homogeneous segments. We aim to develop and expand general change point procedures for FTS. These will importantly make use of global selection criteria tailored to functional data for performing multiple change point analysis, as well as robust approaches to tune these methods. White Noise Tests: A classical problem, crucial in developing models for FTS, is to evaluate whether a time series contains any meaningful "signal", or instead appears to evolve as pure noise. We will extend white noise tests for FTS in several important directions, including the evaluation of the goodness-of-fit of oft-used FTS models. We also develop methods for detecting non-standard signals in FTS not captured by existing approaches. Volatility Models: Models for volatile FTS provide novel ways of incorporating high frequency asset price data to measure financial risk. These models so far rely on key simplifying assumptions that severely limit their flexibility. We will correct these deficiencies by proposing new, more complete, volatility models for FTS, along with accompanying estimation and model identification procedures. Robust Methods: Essentially all existing methods to study FTS are highly susceptible to the influence of outliers or irregularities in the data. We aim to develop companion robust alternatives to existing methods in FTSA, notably to measure autocorrelation and perform dimension reduction tailored to FTS. Applications to Cancer Cell Biology and Immunotherapy: Imaging data from cancer biopsies using modern microscopes capture the motion of immune cells as they interact with cancer cells. Based on these data, we will develop methods in FTSA useful to improve our understanding of interactions between immune cells and cancer cells, and to predict success of emerging therapies.