Entrance hall of the ICMS Workshop

Learning graphical models in high dimensional settings

Apr 04, 2017 - Apr 07, 2017

ICMS, 15 South College Street

Organisers

Name Institution
Evans, Robin University of Oxford
Massa, Sofia University of Oxford
Scutari, Marco University of Oxford

 

The purpose of this workshop is to help meet the theoretical, methodological and applied challenges in extending graphical models to higher dimensional settings. It has two major objectives:

  1. To promote the development of new cutting-edge research at the interface of theory, modelling and applications for multivariate dependences, in order to face the challenges presented when trying to model large and complex data sets.  This will happen through an extensive exchange of ideas between leading researchers who use multivariate modelling in applied contexts on the one hand (for example in biology, genomics, medicine, social science), and leading experts in the theoretical aspects of graphical models on the other.

  2. To stimulate the collaboration between researchers from different institutions, in particular to provide opportunities for young researchers to establish new collaborations and partnerships.

This workshop will provide outstanding training, learning, and collaboration opportunity for all participants, especially for young researchers. The UK community, and particularly students and Postdocs, will specifically benefit from the presence of experts based outside the UK. The main impact will be through the bringing together of disparate strands of the community. Theoreticians will benefit from seeing and discussing the latest inferential problems in applied areas, and new statistical methodology; this will inform their future research directions. Those at the applied end of the spectrum will be able to see the latest developments in theory and computation, and use it to develop new methods for their datasets. We anticipate that the meeting will help to form important new collaborations in this rapidly developing field.

 

Arrangements

This workshop will commence with Registration around 09.00 on Tuesday 4 April and close at lunchtime on Friday 7 April 2017.

 

Participation

  • INVITED SPEAKERS: You will have received an invitation email from ICMS in December 2016 or January 2017.
    - Please register (by 3 February 2017)  via the link in your invitation email.
  • PUBLIC APPLICANTS: THE PUBLIC APPLICATION PERIOD IS NOW CLOSED.
    - Accommodation will be provided, if required.
    - You are expected to arrange and pay for your own travel.
    - You are invited to contribute a poster. Please provide a title and short abstract on the application form (no more than 200 words).
    - As places re limited, please state your research interests on the application form. 


Registration Fee
A registration fee of 100 GBP is payable by all participants (excluding Organisers).  Payment can be made on arrival at ICMS - at Registration we will accept cash, credit/debit card payments and sterling cheques (payable to “Heriot-Watt University”).  Unfortunately, we do not yet have an on-line payment system at ICMS.   If you wish to pay by credit/debit card, please complete this credit/debit card form and bring the completed form to Registration. 

 

Venue
The workshop will be held at ICMS, 15 South College Street, Edinburgh. You may also find this map useful for the workshop.

 

UK Visas
If you are travelling from overseas you may require an entry visa. A European visa does not guarantee entry to the UK. Please use this link to the UK Visas site to find out if you need a visa and if so how to apply for one. 

 

Travel
Information about travel to the UK and Edinburgh is available here.

Please note that it is your responsibility to have adequate travel insurance to cover medical and other emergencies (volcanic ash disruptions) that may occur on your trip.

A taxi directly from the airport will cost approximately 20.00 to 25.00 GBP to the city centre for a one-way journey.     There is also a bus service direct from the airport to the city centre which will cost 4.50 GBP single or 7.50 GBP return - the Airlink 100.  This is a frequent service (every 10 minutes during peak times) and will bring you close to Waverley Railway Station, only a short walk to the accommodation and the workshop venue (see map in 'Venue' section above). 

Lothian buses charge £1.60 for a single, £4.00 for a day ticket. Please note that the exact fare is required and no change is given.

If travelling by train, please note that Edinburgh has several railway stations - Waverley Railway Station being the main station and closest to the workshop venue at 15 South College Street.  If you alight at Edinburgh Waverley, the workshop venue is an easy 10 minute walk over North and South Bridge. The second large railway station is called Haymarket and is at the West End of the city centre.    Please be aware that there is also an Edinburgh Park railway station but this is at the west side of the city some way from the city centre.  For ICMS, use Edinburgh Waverley.

 

Talks and Audio/Visiual Equipment
All talks will be held in the Newhaven Lecture Theatre.  The lecture theatre has a built in computer, data projector, and visualiser/document camera.  In addition, there are two blackboards.  The projector and one board may be used simultaneously.  We advise Speakers that, where possible, you bring your talk on a memory stick/USB to put onto our computer in advance of your session - either at the start of the day or during coffee/lunch breaks.   ICMS staff can help with this.   It is possible for you to use your own laptops but it is then your own responsibility to ensure that resolutions are changed on the laptops if necessary (ICMS will not change the resolution on our equipment).   If you use a MAC we expect you to bring your own adaptor.  

 

Poster Session
There will be a Poster Session on Wednesday, following the afternoon coffee break in the Chapterhouse, Level 1 at ICMS.  Please hand your poster to a member of ICMS staff at Registration and we will display the poster for you in time for the Poster Session.    We can accommodate any size of poster, landscape or portrait, although we'd ask that you limit the poster to a maximum of A0 size to make it easier to display.   If you need to print your poster when you arrive in Edinburgh, there is a "Callprint" printers in the same street as ICMS.   

 

Public Lecture
On Thursday 6 April, Professor Peter Bühlmann will give a Public Lecture.   Doors will open to the public at 17.30, with the lecture starting at 18.00.   The lecture will be followed by an informal wine reception in the Chapterhouse on Level 1.   Seats have been reserved for the workshop participants, so no action is needed to attend.

  • Peter Bühlmann (ETH Zürich)
    Cause or effect? Looking beyond correlation

    Is it a cause or an effect? This simple but fundamental question is often asked in society and science. Searching for the causes for a certain disease and quantifying their effects on the disease status is a long-standing problem in medical studies. Considering correlations among measured variables is insufficient for distinguishing cause from effect. The classical and well-developed framework for causal inference is based on randomised studies; with the drawback that they are often very expensive or even impossible to do due to ethical reasons. Recent approaches try to ''substitute in part'' the randomised studies by models, algorithms and statistical techniques. Although thought-provoking, they can be particularly attractive in nowadays Big Data settings.

 

Catering
The workshop grant will provide refreshments during the breaks in the Programme.  Lunch on Tuesday, Wednesday and Thursday is provided (no lunch is provided on Friday).   During the Poster Session, wine and soft drinks will be available.   The workshop dinner will take place (free of charge to participants) in Blonde Restaurant, 19.00, on Wednesday 5 April at 19.00.  Following the Public Lecture on Thursday 6 April, there will be an informal wine reception.  

 

Wireless Access
The workshop venue, 15 South College Street, has wireless access throughout. EDUROAM works throughout the building.  If you do not use EDUROAM, on arrival at Registration you will be given instructions and a code for accessing another wireless network.  For those without laptops, there will also be a computer available for you to check your emails. 

 

 

Programme

Please note that the following programme is provisional and may yet be subject to alteration.

Download a pdf of the presentation abstracts

Download a pdf of the poster abstracts

Tuesday 4 April

09:30-10:15

Registration in the Chapterhouse, Level 1

10:15-10:30

Opening Session

10:30-11:15

James Cussens (University of York)
Learning graphical models using integer programming
pdf of presentation

11:15-12:00

Natalia Bochkina (University of Edinburgh)
Joint statistical inference for two similar sparse graphical models with application to gene expression data

12:00-13:30

Lunch provided in the Chapterhouse

13:30-14:15

Anna Klimova (National Center for Tumor Diseases, Germany)
Learning tree structures in hematopoiesis
pdf of presentation

14:15-15:00

Marco Scutari (University of Oxford)
Bayesian networks, MAGIC populations and mutliple trait prediction
pdf of presentation

15:00-15:30

Tea/Coffee in the Chapterhouse

15:30-16:05

Christine Sinoquet (University of Nantes)
Advanced graphical models to uncover the genetical factors responsible for a phenotype in genome wide datasets
pdf of presentation

16:05-16:40

Therese Graversen (University of Copenhagen)
An expectation calculator based on probability propagation

16:40-17:15

Gwenael Leday (University of Cambridge)
Bayes factors for associations between variables

17:15-18:30

Informal wine reception in the Chapterhouse

 

Wednesday 5 April

09:30-10:05

Francesco Stingo (University of Florence)
Bayesian graphical models for complex structured data

10:05-10:40

Tamás Rudas (Hungarian Academy of Sciences & Eötvös Loránd University)
On testing graphical models in high dimensional settings

10:40-11:10

Tea/Coffee in the Chapterhouse

11:10-11:45

Marco Scutari (University of Oxford)
Beyond uniform priors in Bayesian network structure learning
pdf of presentation

11:45-12:30

Guido Consonni (Università Cattolica del Sacro Cuore)
Objective Bayes selection of covariate-adjusted Gaussian graphical models

12:30-13:30

Lunch provided in the Chapterhouse

13:30-14:15

Milan Studený (Academy of Sciences of the Czech Republic)
Towards ILP learning decomposable models: chordal graph polytope
pdf of presenation 

14:15-15:00

Sebastian Stolze (University of Bath)
Using Bayesian networks for quantitative risk analysis

15:00-15:30

Tea/Coffee in the Chapterhouse

15:30-17:15

Poster Session

19.00

Workshop dinner at Blonde Restaurant, 75 St Leonard’s Street, Edinburgh

 

Thursday 6 April

09:30-10:05

Marcos Bueno (Radboud University Nijmegen)
Learning asymmetric hidden Markov models

10:05-10:40

Philippe Leray (Université de Nantes)
Learning probabilistic relational models

10:40-11:10

Tea/Coffee in the Chapterhouse

11:10-11:45

Robert Ness (Zymergen, Inc)
A Bayesian active learning experimental design for inferring signaling networks

11:45-12:30

Marco Grzegorczyk (Groningen University)
Advanced dynamic Bayesian network models

12:30-13:30

Lunch provided in the Chapterhouse

13:30-14:15

Jim Smith (University of Warwick)
Selecting BNs with hidden variables

14:15-15:00

Dominik Rothenhäusler (ETH Zürich)
Invariant causal prediction in linear SEMs with latent variables

15:00-15:30

Tea/Coffee in the Chapterhouse

15:30-16:05

Marloes Maathuis (ETH Zürich)
Structure learning in the presence of weak edges

16:05-16:40

Elena Stanghellini (Università Degli Studi di Perugia)
On the identification of discrete graphical models with hidden variables
pdf of presentation

16:40-17:15

Benjamin Frot (ETH Zürich)
Learning directed acyclic graphs in presence of some latent variables

17:30-18:00

Doors open to the public

18:00-19:00

Public Lecture by Peter Bühlmann (ETH Zürich)
Cause or effect? Looking beyond correlation

19:00-20:00

Informal wine reception in the Chapterhouse, Level 1

 

Friday 7 April

09:30-10:05

František Matúš (Institute of Information Theory and Automation)
On patterns of conditional independences and covariance signs among binary variables

10:05-10:40

Kayvan Sadeghi (University of Cambridge)
A constraint-based algorithm for model selection from faithful distributions

10:40-11:10

Tea/Coffee in the Chapterhouse

11:10-11:45

Robin Evans (University of Oxford)
Geometry of graphical model selection

11:45-12:30

Nanny Wermuth (Johannes Gutenberg University Mainz)
Which large Ising models can be studied and generated via simple linear relations?

12:30-12:40

Closing session

 

Download a pdf of the poster abstracts

Download a pdf of the abstracts

 

Presentations:

Presentation Details
Anacleto, Osvaldo
POSTER: Dynamic chain graph models for time series network data
View Abstract Down
Asvatourian, Vahé
POSTER: Estimating causal effects of repeated multi-dimensional biomarkers in observational setting
View Abstract Down
Bochkina, Natalia
Joint statistical inference for two similar sparse graphical models with application to gene expression data
View Abstract Down
Bueno, Marcos
Learning asymmetric hidden Markov models
View Abstract Down
Consonni, Guido
Objective Bayes selection of covariate-adjusted Gaussian graphical models
View Abstract Down
Cussens, James
Learning graphical models using integer programming
View Abstract Down
Didelez, Vanessa
Instrumental variable methods with high-dimensional covariates
View Abstract Down
Djordjilovic, Vera
POSTER: Challenges of mediation analysis in high-dimensional epigenomic studies
View Abstract Down
Doretti, Marco
POSTER: Tackling non-ignorable dropout in the presence of time-varying confounding: a graphical model based approach
View Abstract Down
Evans, Robin
Geometry of graphical model selection
View Abstract Down
Foraita, Ronja
POSTER: Joint author of poster with Moritz Hanke
Frot, Benjamin
Learning directed acyclic graphs in presence of some latent variables
View Abstract Down
Gottard, Anna
POSTER: Graphical models for non-linear and quasi-linear systems
View Abstract Down
Graversen, Therese
An expectation calculator based on probability propagation
View Abstract Down
Grzegorczyk, Marco
Advanced dynamic Bayesian network models
View Abstract Down
Häggström, Jenny
POSTER: Data-driven confounder selection via Markov and Bayesian networks
View Abstract Down
Hanke, Moritz
POSTER: Selecting an optimal tuning parameter based on prior knowledge
View Abstract Down
Klimova, Anna
Learning tree structures in hematopoiesis
View Abstract Down
Leday, Gwenael
Bayes factors for associations between variables
View Abstract Down
Leray, Philippe
Learning probabilistic relational models
View Abstract Down
Maathuis, Marloes
Structure learning in the presence of weak edges
View Abstract Down
Matúš, František
On patterns of conditional independences and covariance signs among binary variables
View Abstract Down
McCann, Mark
POSTER: Pilot protocol for graphical synthesis of high dimensional evidence: diagram based analysis of causal systems (Dacs) as an approach to alcohol harm introduction
View Abstract Down
Ness, Robert
A Bayesian active learning experimental design for inferring signaling networks
View Abstract Down
Peluso, Stefano
POSTER: Objective Bayes selection of essential Gaussian graphical models
View Abstract Down
Rothenhäusler, Dominik
Invariant causal prediction in linear SEMs with latent variables
View Abstract Down
Rudas, Tamás
On testing graphical models in high dimensional settings
View Abstract Down
Sadeghi, Kayvan
A constraint-based algorithm for model selection from faithful distributions
View Abstract Down
Scutari, Marco
TALK 1: Bayesian networks, MAGIC populations and mutliple trait prediction TALK 2: Beyond uniform priors in Bayesian network structure learning
View Abstract Down
Sinoquet, Christine
Advanced graphical models to uncover the genetical factors responsible for a phenotype in genome wide datasets
View Abstract Down
Smith, Jim
Selecting BNs with hidden variables
View Abstract Down
Stanghellini, Elena
On the identification of discrete graphical models with hidden variables
View Abstract Down
Stingo, Francesco
Bayesian graphical models for complex structured data
View Abstract Down
Stolze, Sebastian
Using Bayesian networks for quantitative risk analysis
View Abstract Down
Studený, Milan
Towards ILP learning decomposable models: chordal graph polytope
View Abstract Down
Wang, Yuhao
POSTER: A new framework for joint estimation of multiple high-dimensional directed acyclic graphs
View Abstract Down
Wermuth, Nanny
Which large Ising models can be studied and generated via simple linear relations?
View Abstract Down

Participants

Name Institution
Anacleto, Osvaldo University of Edinburgh, Roslin Institute
Asvatourian, Vahé Gustave Roussy Institute
Bochkina, Natalia University of Edinburgh
Bueno, Marcos Radboud University Nijmegen
Bühlmann, Peter ETH Zürich
Consonni, Guido Universita' Cattolica del Sacro Cuore
Cussens, James University of York
Djordjilovic, Vera University of Oslo
Doretti, Marco University of Perugia
Evans, Robin University of Oxford
Foraita, Ronja Leibniz Institute for Prevention Research and Epidemiology - BIPS
Frot, Benjamin ETH Zürich
Gottard, Anna DiSIA - University of Florence
Graversen, Therese University of Copenhagen
Grzegorczyk, Marco Groningen University
Häggström, Jenny Umeå University
Hanke, Moritz Leibniz Institute for Prevention Research and Epidemiology - BIPS
Højsgaard, Søren Aalborg University
Klimova, Anna National Center for Tumor Diseases, Germany
Leday, Gwenael University of Cambridge
Leray, Philippe Université de Nantes
Maathuis, Marloes ETH Zürich
Massa, Sofia University of Oxford
Matúš, František Institute of Information Theory and Automation
McCann, Mark University of Glasgow
Ness, Robert Zymergen, Inc
Peluso, Stefano Università Cattolica del Sacro Cuore
Rothenhäusler, Dominik ETH Zürich
Rudas, Tamás Hungarian Academy of Sciences & Eötvös Loránd University
Sadeghi, Kayvan University of Cambridge
Scutari, Marco University of Oxford
Sinoquet, Christine University of Nantes
Smith, Jim University of Warwick
Stanghellini, Elena Università Degli Studi di Perugia
Stingo, Francesco University of Florence
Stolze, Sebastian University of Bath
Studený, Milan Academy of Sciences of the Czech Republic
Wang, Yuhao Massachusetts Institute of Technology
Wermuth, Nanny Johannes Gutenberg University Mainz