Statistics Across Campuses

Changes in rainfall and flooding across Australia Date: 24 June 2022, Friday Time: 4pm AEDT Speaker: Dr Conrad Wasko  (University of Melbourne) Abstract:  As the climate warms extreme rainfall should increase as dictated by the Clausius-Clapeyron relationship. Knowing that a thermodynamic relationship exists linking rainfall and temperature it could be expected that historical sensitivities of rainfall and temperature would be useful in informing us how climate change will affect rainfall. But in practice the estimation sensitivities are fraught due to artefacts that arise in their calculation. Despite this, there is evidence to suggest they may be helpful in informing potential changes in rainfall and flooding in a warmer world.   As changes in rainfall extremes are expected to increase flooding it is critical to consider changes in flooding for infrastructure design. The second part of this talk focusses on the application of circular statistics to understanding changes in flood timi

  A general-purpose method for supervised learning under covariate shift  with applications to observational cosmology. Date: Tuesday, 14 June 2022   Time: 16:00pm (UTC)   Speaker:  Prof. Roberto Trotta (SISSA, Italy) Contact the organizer: Andriy Olenko a.olenko@latrobe.edu.au Abstract: Supervised machine learning will be central in the analysis of upcoming large-scale sky surveys. However, selection bias for astronomical objects yields labelled training data that are not representative of the unlabelled target data distribution. This affects the predictive performance with unreliable target predictions and poor generalization. I will present StratLearn, a novel and statistically principled method to improve supervised learning under such covariate shift conditions, based on propensity score stratification. In StratLearn, learners are trained on subgroups ("strata") of the data conditional on the propensity scores, leading to improved covariate balance and much-reduced bias

The Annealed Leap-Point MCMC Sampler (ALPS) for multi-modal posterior distributions Date: 10 June 2022, Friday Time: 4pm AEDT Speaker:  Dr Matt Moores  (University of Wollongong) Abstract: Multi-modal distributions pose major challenges for the usual algorithms that are employed in statistical inference. These problems are exacerbated in high-dimensional settings, where techniques such as Markov Chain Monte Carlo (MCMC) and Expectation Maximisation (EM) typically rely upon localised update mechanisms: such localised algorithms can effectively become trapped in one of the local modes, leading to biased inference and underestimation of uncertainty.   In this talk, I will introduce the Annealed Leap-Point Sampler (ALPS), an MCMC algorithm that augments the state space of the target distribution with a sequence of modified, annealed (cooled) distributions. The temperature of the coldest state is chosen such that the corresponding annealed target density of each individual mode can be close