Friday, 30 November, 2018
The University of Queensland, Emmanuel College, Riverview Room
Monte Carlo methods permeate much of contemporary science, data science, and machine learning. This workshop brings together practitioners and theorists in Monte Carlo simulation with the aim to highlight new theoretical developments and focus on new challenges ahead.
Sandeep Juneja, Tata Institute of Fundamental Research: Sample complexity of partition identification using multi-armed bandits with applications to nested Monte Carlo
- Fred Roosta, The University of Queensland: How Randomized Analysis Can Help Us Do “Data Science:” Examples in Deep Learning and Graph Analysis
- Chris Drovandi, Queensland University of Technology: Sequential Monte Carlo for Static Bayesian Models
- Xuhui Fan, University of New South Wales: Rectangular Bounding Process
- Matias Quiroz, University of New South Wales: The block-Poisson estimator for optimally tuned signed pseudo-marginal MCMC
- Sarat Moka, The University of Queensland: Perfect Sampling for Gibbs point processes using partial rejection sampling
- Rob Salomone, The University of Queensland: An introduction to Stein kernels.
- Tui Nolan, University of Technology Sydney: Variational Bayesian Methods for Multilevel Data Models
- Gael Martin, Monash University: Loss-based Bayesian Prediction
- Edwin Bonilla, CSIRO Data 61: Variational Network Inference- Strong and Stable with Concrete Support
Organisers: Robert Kohn, UNSW and Dirk Kroese, UQ