Graphical Models: Conditional Independence and Algebraic Structures

October 23-25, 2019

The goal of this workshop is to bring together researchers in graphical models and explore the strong algebraic connections present in their analysis, with a focus on conditional independence.

SPEAKERS

Carlos Améndola Abstract
James Cussens Abstract
Vanessa Didelez Abstract
Mathias Drton   Abstract
Eliana Duarte Abstract
Robin Evans Abstract
Dominik Janzing Abstract
Thomas Kahle   Abstract
Claudia Klüppelberg   Abstract
Kaie Kubjas Abstract
Steffen Lauritzen Abstract
Manuele Leonelli   Abstract
Marloes Maathuis Abstract
Bartosz Kolodziejek Abstract
Johannes Rauh Abstract
Elina Robeva  Abstract
Liam Solus Abstract
Seth Sullivant Abstract
Henry Wynn Abstract
Piotr Zwiernik Abstract

CONFERENCE PROGRAMME

The workshop will begin on Wednesday around 9 am and end on Friday at 12:30 pm.

08:30 - 09:00       Registration
09:00 - 09:10       Welcome Words
09:10 - 09:55       Carlos Améndola
                                Introduction to max-linear models and tropical linear algebra
09:55 - 10:40       Steffen Lauritzen
                                Conditional independence of max-linear graphical models
10:40 - 11:10       Coffee Break
11:10 - 11:55       Eliana Duarte
                                Gröbner bases for staged tree models
11:55 - 12:40       Thomas Kahle
                                The geometry of gaussoids
12:40 - 14:00       Lunch Break
14:00 - 14:45       Dominik Janzing
  Non-statistical notions of independence in causal discovery
14:45 - 15:30       Bartosz Kolodziejek
                                Model selection in the class of Gaussian models invariant under a subgroup of the symmetric group
15:30 - 16:00       Coffee Break
16:00 - 16:45       Piotr Zwiernik
                                Learning the structure of graphical models by covariance queries
16:45 - 17:30       Claudia Klüppelberg
                              Identifiability and estimation of recursive max-linear models
17:30 - 19:30       Poster Session

_____________________

09:00 - 09:45       Marloes Maathuis
  False discovery rate control for Gaussian graphical models
09:45 - 10:30       Robin Evans
                                Angles and model selection
10:30 - 11:00       Coffee Break
11:00 - 11:45       Kaie Kubjas
                          Maximum likelihood estimation of toric fiber products
11:45 - 12:30       Johannes Rauh
                                Marginal faces of marginal polytopes
12:30 - 14:00       Lunch Break
14:00 - 14:45       James Cussens
                                On the polyhedral geometry of conditional independence relations
14:45 - 15:30      Manuele Leonelli
                              Model-preserving sensitivity analysis for families of Gaussian distributions
15:30 - 16:00       Coffee Break
16:00 - 16:45       Liam Solus
                                Interventional Markov equivalence for mixed graph models
16:45 - 17:30       Vanessa Didelez
                                Assumptions for Mendelian randomization studies with multiple instruments
18:00 - 23:00       Conference Dinner

____________________

09:00 - 09:45 Seth Sullivant
                                 Ideals of Gaussian graphical models
09:45 - 10:30 Elina Robeva
                                Nested covariance determinants in Gaussian graphical models
10:30 - 11:00 Coffee Break
11:00 - 11:45 Henry Wynn
                                Lattice conditional independence models and Hibi rings
11:45 - 12:30 Mathias Drton
                                Causal discovery in linear non-Gaussian models
12:30              Closing

CONFERENCE SLIDES

Carlos Améndola: Introduction to max-linear models and tropical linear algebra
James Cussens: On the polyhedral geometry of conditional independence relations
Mathias Drton: Causal discovery in linear non-Gaussian models
Eliana Duarte: Gröbner bases for staged tree models
Robin Evans: Angles and Model Selection
Dominik Janzing: Non-statistical notions of independence in causal discovery
Thomas Kahle: The geometry of gaussoids
Claudia Klüppelberg: Identifiability and estiation of recursive max-linear models
Bartosz Kołodziejek: Model selection in the class of Gaussian models invariant under a subgroup of the symmetric group
Steffen Lauritzen: Conditional independence in max-linear Bayesian networks
Manuele Leonelli: Model-preserving sensitivity analysis for families of Gaussiandistributions
Johannes Rauh: Marginal faces of marginal polytopes
Elina Robeva: Nested Covariance Determinants in Gaussian Graphical Models
Seth Sullivant: Ideals of Gaussian Graphical Models
Piotr Zwiernik: Learning the structure of graphical models by covariance queries

CONFERENCE LOCATION

Institute for Advanced Study (IAS)
Technische Universität München
Lichtenbergstraße 2 a
85748 Garching b. München

CONFERENCE DINNER

On Thursday, October 24 there will be a joint dinner at the Garchinger Augustiner. Please confirm your attendance by emailing grant@ma.tum.de until October 10, 2019. You can mention in your email If you need a receipt or certificate of attendance to the workshop.

ACCOMMODATION

Munich is a popular travel destination so we suggest that you book your accommodation early.

All speakers, as well as participants who requested so, have been booked at Motel One München-Garching.

SCIENTIFIC ORGANIZERS

Carlos Améndola, Claudia Klüppelberg, Steffen Lauritzen, and Ngoc Tran