The next fortnightly BRAG meeting will be held on Thursday the 4th of April at 1 pm via Zoom/GP-Y801. For this session we will have a workshop-style sequential Monte Carlo (SMC) themed BRAG. Adam Bretherton will provide an introduction to SMC, and Sarah Vollert will outline some applications of SMC from her research, with some time for questions and discussion.
Zoom Link: https://qut.zoom.us/j/84965931799?pwd=ajFUZUUxSXdXRUZqTkFXWTd2MVBRdz09
Password: 731848
An introduction to sequential Monte Carlo (SMC) – Adam Bretherton
Sequential Monte Carlo (SMC) algorithms are a set of simulation based methods for approximating the posterior distribution of interest. These methods build on adaptive importance sampling by moving a population of particles through a sequence of distributions, typically from the prior and finishing at the target posterior. In this talk we provide a brief introduction to importance sampling, SMC and adaptive SMC.
SMC in practice – Sarah Vollert Sequential Monte Carlo algorithms are typically a good choice if there is a low probability of randomly sampling good parameter values, or if there is no well-defined likelihood peak – both of which are common issues in ecological and biological applications. Here, I show the power of approximate Sequential Monte Carlo algorithms for parameterisation in ecological conservation modelling and in cellular biology.
Bayesian Research & Applications Group 