Workshop overview
The computational statistics revolution of the 1990s provided powerful methodology for carrying out likelihood-based inference, including Markov chain Monte Carlo methods, the EM algorithm, many associated optimisation techniques for likelihoods, and Sequential Monte Carlo methods.

Although these methods have been and are highly successful in making likelihood-based inference accessible to a wide range of problems, we now have a far better understanding of their limitations, for example in high-dimensional problems and for massive data sets. Thus many challenging statistical inference problems of the 21st century cannot be addressed using existing likelihood-based methods. Motivating examples come from genetics, genomics, infectious disease epidemiology, ecology, commerce, and bibliometrics.

However recent breakthroughs in computational and statistical approaches hold promise to address intractable likelihood problems, including pseudo-marginal and particle MCMC, likelihood-free methods such as Approximate Bayesian Computation, composite and pseudo-likelihoods, new simulation methods for hitherto intractable stochastic models, and adaptive Monte Carlo methods. These advances coupled with developments in multi-core computational technologies such as GPUs, have enormous potential for extending likelihood methods to meet the most difficult challenges of modern scientific questions.

This workshop will bring together leaders in the field of computational statistics and scientific computing to discuss their latest ideas and present methods for addressing complex inference problems.

Draft timetable

The venue for the workshop is the Tsuzuki Lecture Theatre at St Anne's College.
Thursday 20th March
10:00 — 10:30 Registration
10:30 — 11:15 Christophe Andrieu
11:15 — 12:00 Chris Sherlock
12:00 — 13:30 Lunch
13:30 — 14:15 Mike Giles
14:15 — 15:00 Nick Whiteley
15:00 — 15:45 Tea
15:45 — 16:30 Helen Ogden
16:30 — 17:15 Jim Griffin
17:15 — 19:00 Posters and drinks
19:00 -- Dinner
        
Friday 21st March
09:15 — 10:00 Paul Fearnhead
10:00 — 10:45 Anthony Lee
10:45 — 11:05 Tea
11:05 — 11:50 Sumeet Singh
11:50 — 12:35 Krys Łatuszyński
12:35 — 13:45 Lunch
13:45 — 14:30 Rémi Bardenet
14:30 — 15:15 Éric Moulines
15:15 Finish
The poster session in the evening on Thursday will take place in Foyer A of the Ruth Deech Building.

Speakers

The speakers at the workshop will include (in no particular order):
  • Éric Moulines (Institut Télécom-Mines / Télécom ParisTech)
     ‘Metropolis adjusted proximal gradient algorithm’
  • Mike Giles (University of Oxford)
     ‘Trends in HPC: hardware complexity and software challenges’
  • Rémi Bardenet (University of Oxford)
     ‘Scaling up MCMC: a subsampling approach’
  • Jim Griffin (University of Kent)
     ‘An adaptive MCMC scheme for variable selection problems’
  • Christophe Andrieu (University of Bristol)
     ‘Establishing some order amongst exact approximation MCMCs’
  • Nick Whiteley (University of Bristol)
     ‘Particle filtering under interaction constraints’
  • Sumeet Singh (University of Cambridge)
     ‘Asymptotic behaviour of approximate Bayesian estimators’
  • Krys Łatuszyński (University of Warwick)
     ‘The Bernoulli factory, extensions & applications’
  • Anthony Lee (University of Warwick)
     ‘Variance bounding & geometric ergodicity of Markov chain Monte Carlo kernels for approximate Bayesian computation’
  • Helen Ogden (University of Warwick)
     ‘Inference for generalized linear mixed models with sparse structure’
  • Chris Sherlock (Lancaster University)
     ‘Optimising pseudo-marginal random walk Metropolis algorithms’
  • Paul Fearnhead (Lancaster University)
     ‘Reparameterisations for Particle MCMC’

Posters

If you would like to present a poster, please send a title and brief abstract to mitchell@stats.ox.ac.uk stating it is a poster for the i-like workshop.

Registration fees

  • £90 for students (undergraduate, MSc or PhD)
  • £135 others
The fee includes tea/coffee, lunch on Thursday and Friday, and workshop dinner in St Anne’s on Thursday evening.
Note that registration does not cover accommodation.

Please note due to space constraints, registration is limited to 50 people excluding speakers and organising committee.

Registration is now open via the University of Oxford conference booking system.

Accommodation

Although registration does not cover accommodation, St Anne's College are providing a limited number of rooms which can be booked directly online here using the promotional code STATS17837 for £90 (inc. VAT) per night.

Breakfast is included.  Check-in is after 1 p.m. and check out is no later than 10 a.m. on the day of departure.

Travel

The workshop will take place in St Anne's College, located here.

Rail
Oxford has a mainline railway station which is a 20 minute walk from St Anne's College, or 5 minutes in a taxi.

Coach/Bus
The coach station at Gloucester Green is well served by Stagecoach, National Express and the Oxford Bus Company.  It is a 12 minute walk from St Anne's College, or a few minutes by taxi.

Driving
Oxford is close to both the M40 (junction 8A or 9) and a little further from the M4 (junction 13 or 15).  However, parking in Oxford is very difficult and St Anne's itself has very little parking for those taking accommodation there.  There are good park & ride facilities on the outskirts, with Pear Tree park and ride having stops a short walk from St Anne's College.  Pear Tree costs £2 per day, and the journey in on the bus takes approximately 20 minutes and has frequent departures most of the day.

By Air
The airports which are easiest to travel to are Heathrow and Gatwick by coach using The Airline coaches, or Birmingham by mainline train (Birmingham International station, which is linked to the airport by light rail shuttle).