Mathematics Colloquium: Matias Quiroz
Topic: Subsampling MCMC: Bayesian inference for large data problems
The rapid development of computing power and efficient Markov chain Monte Carlo (MCMC)
simulation algorithms have revolutionized Bayesian statistics, making it a highly practical
inference method in applied work. However, MCMC algorithms tend to be computationally
demanding, and are particularly slow for large datasets. In this talk I will present Subsampling
MCMC, a so called pseudo-marginal MCMC approach to speeding up MCMC through data
subsampling. I will cover both an approximate and an exact approach, and show how to extend
them to high-dimensional problems using Hamiltonian Monte Carlo.
About the speaker
Matias Quiroz received his Ph.D. in Statistics from Stockholm University in 2015. Matias joined the
University of New South Wales as a Postdoctoral fellow during 2017, where he works under the
supervision of Professor Robert Kohn. Matias is a Bayesian believer and works on Bayesian
computation, in particular Markov chain Monte Carlo and more recently variational Bayes.