OxCSML seminar this Friday 10 March
26 views
Skip to first unread message
Hai Dang Dau
Mar 7, 2023, 9:17:50 AM3/7/23
to oxcsml...@googlegroups.com, oxc...@googlegroups.com
Dear all,
Please find below the detail of the OxCSML seminar this week. Hope to see many of you there.
Best,
Hai-Dang.
==================
Time and place: 16.00, Friday 10 March, Large Lecture Theatre, Department of Statistics, University of Oxford
Speaker: Tui Nolan, MRC Biostatistics Unit, University of Cambridge
Title: Orthogonal Bayesian Functional Principal Components Analysis
Abstract: Modelling of longitudinal datasets is an important methodological task in various fields of applied science. In this talk, we will focus on daily weather data collected from 2837 weather stations within the United States of America from 1960–1994. In this particular example, we have two levels of longitudinal data: the first level is the yearly temperature recordings, and the second level is the variation in temperature across 1960–1994. The large number of weather stations and the 45 years of daily measurements results in an extremely large modelling task. An efficient way to deal with such problems is through functional principal components analysis (FPCA), and in this particular example, we will focus on multilevel FPCA. Standard approaches for FPCA rely on an eigendecomposition of a smoothed covariance surface in order to extract the orthonormal eigenfunctions representing the major modes of variation in a set of functional data. This approach can be a computationally intensive procedure, especially in the presence of large datasets with irregular observations. In this article, we develop a variational Bayesian approach, which aims to determine the Karhunen-Lo`eve decomposition directly without smoothing and estimating a covariance surface. More specifically, we incorporate the notion of variational message passing over a factor graph because it removes the need for rederiving approximate posterior density functions if there is a change in the model. Instead, model changes are handled by changing specific computational units, known as fragments, within the factor graph. Indeed, this is the first article to address a functional data model via variational message passing. Our approach introduces two new fragments that are necessary for Bayesian functional principal compo- nents analysis. We present the computational details, a set of simulations for assessing the accuracy and speed of the variational message passing algorithm and an application to United States temperature data.
Speaker: Tui Nolan, MRC Biostatistics Unit, University of Cambridge
Title: Orthogonal Bayesian Functional Principal Components Analysis
Abstract: Modelling of longitudinal datasets is an important methodological task in various fields of applied science. In this talk, we will focus on daily weather data collected from 2837 weather stations within the United States of America from 1960–1994. In this particular example, we have two levels of longitudinal data: the first level is the yearly temperature recordings, and the second level is the variation in temperature across 1960–1994. The large number of weather stations and the 45 years of daily measurements results in an extremely large modelling task. An efficient way to deal with such problems is through functional principal components analysis (FPCA), and in this particular example, we will focus on multilevel FPCA. Standard approaches for FPCA rely on an eigendecomposition of a smoothed covariance surface in order to extract the orthonormal eigenfunctions representing the major modes of variation in a set of functional data. This approach can be a computationally intensive procedure, especially in the presence of large datasets with irregular observations. In this article, we develop a variational Bayesian approach, which aims to determine the Karhunen-Lo`eve decomposition directly without smoothing and estimating a covariance surface. More specifically, we incorporate the notion of variational message passing over a factor graph because it removes the need for rederiving approximate posterior density functions if there is a change in the model. Instead, model changes are handled by changing specific computational units, known as fragments, within the factor graph. Indeed, this is the first article to address a functional data model via variational message passing. Our approach introduces two new fragments that are necessary for Bayesian functional principal compo- nents analysis. We present the computational details, a set of simulations for assessing the accuracy and speed of the variational message passing algorithm and an application to United States temperature data.
Reply all
Reply to author
Forward
0 new messages