Approximate Bayesian Computation (ABC) performs inference when evaluating the likelihood function is infeasible but sampling from the model is possible. First developed in population genetics, ABC is now a well established technique in this and other diverse disciplines including infectious disease epidemiology, astrophysics and econometrics.
The ABC in Edinburgh workshop will feature invited and contributed talks, and will have a poster session to further exchange and discuss recent work on the theoretical underpinnings, computational advances, and applications of approximate Bayesian computation.
If you would like to present your work as a poster, please send us your abstract by June 17 (1 week before the workshop) to abcinedinburgh@gmail.com in plain text.
We use the same poster boards as the main conference. Poster format is A0 portrait (height: 118 cm / 47 in, width: 84 cm / 33 in).
The workshop takes place in Room G07 in the Informatics Forum, opposite Appleton Tower. For further information on travel, lodging and Edinburgh in general, please see here.
Detailed programme with talk abstracts is here. Folder with all presentations is here.
9:00 Tutorial on ABC
9:30 Gael Martin: Approximate Bayesian Forecasting
10:10 Guillaume Kon Kam King: Bayesian nonparametric modelling of trended functional time series
10:30 Coffee + posters
11:00 Alexander Buchholz: Improving approximate Bayesian computation via quasi-Monte Carlo
11:40 TJ McKinley: Approximate Bayesian Computation and History Matching for complex stochastic epidemic models
12:00 Marko Järvenpää: Efficient acquisition rules for model-based approximate Bayesian computation
12:20 Lunch + posters (90min)
13:50 Emille Ishida: ABC in Astronomy
14:30 Jessica Cisewski: Inference for the stellar initial mass function via approximate Bayesian computation
15:10 Matt Moores: Parametric Functional Approximate Bayesian (PFAB) Algorithm for the Potts Model
15:30 Coffee + posters
16:10 Christopher Drovandi: Recent Advances in Bayesian Synthetic Likelihood
16:50 Closing
Registration is done via the ISBA website. Our workshop is part of the "short courses". ISBA charges a fee of 70 US Dollar for attendance. You won't need to register for the full ISBA meeting; registration for the short courses only is possible too.
Jointly organised by Dennis Prangle, Newcastle University, and Michael Gutmann, University of Edinburgh.
You can reach us at abcinedinburgh@gmail.com or our personal email addresses.
The workshop is part of a series of workshops on approximate Bayesian computation (ABC) and likelihood-free inference that includes
ABC in Banff (2017) - ABC in Helsinki (2016) - ABC in Montreal (2015, 2014) - ABC in Sydney (2014) - ABC in Rome (2013) - ABC in London (2011) - ABC in Paris (2009)