BIRS Workshop 17w5025, Feb 19-24, 2017

Validating and Expanding Approximate Bayesian Computation

Organisers: Luke Bornn - Jukka Corander - Gael Martin - Dennis Prangle - Christian Robert - Richard Wilkinson


We would like to opt for a low intensity programme, with a few talks each day and plenty of free or loose time for collaborations and discussions. This would mean a maximum of six talks a day from Monday to Thursday, and possibly one or two talks on the Friday morning. We also thought that having a slot for snapshot presentations for anyone interested, on the first days would bring everyone together and induce discussions and potential collaborations. Each topic in the following program will be attached with a discussion group, towards presenting criticisms and challenges for the subsequent days. All suggestions and criticisms are welcome.

Tentative schedule

Sunday Feb 19 [evening only]

Evening drinks and dinner

Monday Feb 20

Plenary talks and speed-presentations

Tuesday Feb 21

Plenary talks, working groups, debates

Wednesday Feb 22

Plenary talks, debates, free afternoon

Thursday Feb 23

Plenary talks, working groups, debates

Friday Feb 24 [morning only]

Plenary talks, working groups, conclusions

Participants are expected to arrive on Feb 19. (In case of earlier arrivals, please contact BIRS or the Banff Centre for separate accommodation.)

Saturday Feb 18

17:00-19:00 Drinks and/or dinner at the Maclab Bistro for early arrivals (alternatives abound in the town of Banff, easily accessible by foot, short of an ongoing snowstorm)

Sunday Feb 19

17:00-18:00 Drinks in the common room of Corbett Hall

18:00-19:00 Dinner at the Vistas Dinning Room

Monday Feb 20

9:00-10.00 Main topic 1: Convergence/validation of ABC (Paul Fearnhead)

10.00-10:30 Anthony Lee

10:30-11:00 coffee break  [planning for subsequent days]

11:00-12:00 speed presentations

12:00-13:00 lunch

13:00-13:30 Tour of the Banff Centre [optional]

13:30-14:00 Open time

14:00-14:10 Group photo [possible common denominator, like black hat or white tee-shirt???]

14:10-15:10 Main topic 2: Alternatives to ABC (Michael Guttman)

15:10-15:30 tea break  [planning for subsequent days]

15:30-16:30 Main topic 3: High Dimensional ABC (Richard Everitt)

16:30-17:00 Matt Graham

17:00-17:30 further speed presentations

17:30-18:00 Drinks in Corbett Hall

18:00-19:00 dinner

Tuesday Feb 21

9:00-10.00 Main topic 4: Challenging applications (Jukka Corander)

10:00-10:30 Ewan Cameron

10:30-11:00 coffee break  [planning for subsequent days]

11:00-12:00 Discussion on convergence (Judith Rousseau)

12:00-13:30 lunch

14:-00-15:00 Open time [or additional talks]

15:00-15:30 tea break [planning for subsequent days]

15:30-16:30 Main topic 5: ABC and model assessment (David Frazier)

16:30-17:00 Clara Grazian

17:00-18:00 Discussion on alternatives (Jean-Michel Marin)

18:00-19:00 dinner

Wednesday Feb 22

9:00-10.00 Main topic 6:  Dynamic models (Gael Martin)

10:00-10:30 Robert Kohn

10:30-11:00 coffee break [planning for subsequent days]

11:00-11:30 Pierre Jacob

11:30-12:00 Vikash K. Mansinghka

12:00-13:30 lunch

Free afternoon (work or play)

18:00-19:00 dinner

Thursday Feb 23

Day mostly left open for specific papers that participants have offered to present, plus more discussion groups if active.

9:00-9:30 Kyle Cranmer

9:30-10:00 Patrick Muchmore

10:00-10:30 Samuel Kaski

10:30-11:00 coffee break

11:00-12:00 Discussion on dynamic models (Veronika Czellar)

12:00-13:30 lunch

13:30-15:00 free time [or impromptu talks, depending]

15:00-15:30 tea break

15:30-16:00 Dennis Prangle

16:00-16:30 Marco Cusumano-Towner

16:30--18:00 open space for discussion, further talks, group work

18:00-19:00 dinner

Friday Feb 24 

9:00-10.30 Discussions & Talks & Conclusions

10:30-11:00 coffee break

11:00-12:00 Open time for remainers [many participants will leave by a shuttle to the airport around 11am]

12:00-13:30 lunch

Participants are expected to leave on Feb 24. (In case of later departures, please contact BIRS or the Banf Centre for additional accommodation.)