We are happy to announce a seminar by A. Prof Minh-Ngoc Tran and Jamie Gabor. We hope to see you there.
Any questions, please feel free to contact: secretary.nswbranch@statsoc.org.au.
Date: Wednesday, 17th September 2025
Time:
5.00pm - 6.00pm: Minh-Ngoc Talk
6.00pm - 6.30pm: Break and refreshments
6.30pm - 7.00pm: Jamie Gabor Talk
Venue: F07.02.273. Carslaw Building. Carslaw Lecture Theatre 273, University of Sydney and Zoom
RSVP: Register at https://www.statsoc.org.au/event-6318746
Speaker: Minh-Ngoc Tran, Associate Professor, University of Sydney Business School
Title: Natural Gradient Variational Bayes Without Fisher Matrix Analytic Calculations and Its Inversion
Abstract:
We introduce a method for efficiently approximating the inverse of the Fisher information matrix, a crucial step in achieving effective variational Bayes inference. A notable aspect of our approach is the avoidance of analytically computing the Fisher information matrix and its explicit inversion. Instead, we introduce an iterative procedure for generating a sequence of matrices that converge to the inverse of Fisher information. The natural gradient variational Bayes algorithm without analytic expression of the Fisher matrix and its inversion is provably convergent and achieves a convergence rate of order O(log s/s), with s the number of iterations. We also obtain a central limit theorem for the iterates. Implementation of our method does not require storage of large matrices, and achieves a linear complexity in the number of variational parameters. Our algorithm exhibits versatility, making it applicable across a diverse array of variational Bayes domains, including Gaussian approximation and normalizing flow Variational Bayes. We offer a range of numerical examples to demonstrate the efficiency and reliability of the proposed variational Bayes method. Joint work with A. Godichon-Baggioni and D. Nguyen.
Biography:
Minh-Ngoc's main research interests lie in Bayesian methodology and statistical machine learning. He specialises in fast Variational Bayes and simulation-based methods, such as important sampling and sequential Monte Carlo, for estimating complex models with Big Data, and in Lasso-type variable selection methods.
Speaker: Jamie Gabor, PhD Student, School of Mathematics and Statistics, University of Sydney
Title: Novel Bayesian algorithms for ARFIMA long-memory processes: a comparison between MCMC and ABC approaches
Abstract:
We introduce a comparative study of two Bayesian approaches—Markov Chain Monte Carlo (MCMC) and Approximate Bayesian Computation (ABC)—for estimating the parameters of autoregressive fractionally-integrated moving average (ARFIMA) models, which are widely used to capture long-memory in time series data. We propose a novel MCMC algorithm that filters the time series into distinct long-memory and ARMA components, and benchmarked it against standard approaches. Additionally, a new ABC method is proposed, using three different summary statistics used for posterior estimation. The methods are implemented and evaluated through an extensive simulation study, as well as applied to a real-world financial dataset, specifically the quarterly U.S. Gross National Product (GNP) series. The results demonstrate the effectiveness of the Bayesian methods in estimating long-memory and short-memory parameters, with the filtered MCMC showing superior performance in various metrics. This study enhances our understanding of Bayesian techniques in ARFIMA modelling, providing insights into their advantages and limitations when applied to complex time series data.
Biography:
Jamie is a second year PhD student at the University of Sydney under Clara Grazian. His work is on Bayesian parametric and non-parametric methodologies for estimating long-memory models.
@StatSocAus
