Telecon seminar

This group meets every Wednesday at 7 am PT (see time in your timezone), by Skype.
Format: One (or several) paper(s) READ IN ADVANCE by all the participants are discussed. A moderator summarizes them for 20 minutes and then leads a discussion for 40 minutes. To join the group, send email to events @ chalearn . org.

Upcoming seminars

WINTER/SPRING 2017

 Date
2017
 Time  Speaker/
Moderator
 Paper(s) discussed  Slides (optional)  Notes/ recording
TBA  7 am PT
(16h CET)
David Lopez-Paz

Revisiting Classifier Two-Sample Tests

https://arxiv.org/abs/1610.06545


 TBA 7 am PT  Elias Bareinboim  http://causalfusion.ddns.net:4764/
   7 am PT  
 

 7 am PT  
 

7 am PT

 

 7 am PT  

 


SUMMER/FALL 2016 paper list
Causality (with Krikamol, Elias, David, Kun, Jonas, Alexander, Michele, Diviyan, Berna, Louis (louis.wehenkel), Laurine (duchesne.laurine), Lambert (lpatron5), Isabelle,Antoine,Benjamin, Gilles(gilles.blondel) Louis (nyc65225), Olivier (olivier.goudet_1), Girisha (girishagarg).

 Date  Time  Speaker/
Moderator
 Paper(s) discussed  Slides (optional)  Notes/ recording
 Thursday, Jul 14, 2016 8 PM PT Krikamol Muandet (1) The Randomized Causation Coefficient
(2) Towards a Learning Theory of Cause-Effect Inference

paper list
Wednesday, Jul 20, 2016 7 AM PT David Lopez-Paz same +
(3) Discovering Causal Signals in Images
datasets CE challenge
[view][ppt]
notes
Wednesday Jul 27, 2016 7 AM PT Kun Zhang (1) Domain adaptation under target and conditional shift
(2) Multi-Source Domain Adaptation: A Causal View
(3) Domain Adaptation with Conditional Transferable Components
slides :
View Download
notes
Wednesday Sept 7, 2016 7 AM PT Berna Batu Paper preprint: A non-parametric algorithm for discovering triggering patterns of spatio-temporal event types  View Download slides :
View Download
Thesis: Nonparametric Approaches for DiscoveringTriggering Events from Spatio-Temporal Patterns

Wednesday Sept 14, 2016 7 AM PT Isabelle Guyon  Dataset of the cause-effect pair challenge (+ discussion on creating new benchmarks) datasets CE challenge
[view][ppt][details on data]
 challenge website
[discussion summary]


 Wednesday Sept 21, 2016  7 AM PT  Kun Zhang  Discovery and visualization of nonstationary causal models  

http://arxiv.org/abs/1509.08056

 [notes]
 Wednesday Sept 28, 2016  7 AM PT  Krikamol Muandet  Counterfactual Reasoning and Learning Systems: The Example of Computational Advertising (Bottou et al) View  [notes]
 Wednesday Oct 5, 2016  7 AM PT Kun Zhang Discovery and visualization of nonstationary causal models  [continued + discussion on organizing a causality in time series challenge]
View Download
  [notes]
 Wednesday Oct 12, 2016  7 AM PT   Louis Wehenkel   Bottou's paper (continued): RL aspects and gradient policy learning    View Download
  Wednesday Oct 19, 2016  7 AM PT  Elias Bareinboim  Causal Inference and data-fusion  View  
 Wednesday Oct 26, 2016  7 AM PT  Elias Bareinboim  Causal Inference and data-fusion (continued)    
 Wednesday Nov 2, 2016 8 AM PT, 16h Paris  Louis Wehenkel  Opportunities of applications in power systems  http://www.montefiore.ulg.ac.be/~lwh/Presentations/NRC-2015-LW.pdf  
 Wednesday Nov 16, 2016  7 AM PT

 Diviyan Kalainathan et Olivier Goudet   Happiness in the workplace: opportunity to study causal mechanisms Slides View Download  


WINTER/SPRING 2016

 Date  Time  Speaker/
Moderator
 Paper(s) discussed  Slides (optional)  Notes/ recording
 Thursday, Mar 10, 2016 8 am PT



 Thursday, Mar 17, 2016 8 am PT Danny Silver Life Long Learning
meeting minutes
 Thursday, Mar 24, 2016 8 am PT


 
 Thursday, Mar 31, 2016 8 am PT (17h) Sergio Escalera LAP challenges [slides]  Material from conferences I
 Material from conferences II

 Thursday April 7, 2016 8 am PT
 
 
 Thursday, April 14, 2016 8 am PT Isabelle Guyon and Balazs Kegl Challenge platforms. RAMP. Codalab tutorial https://drive.google.com/open?id=0BzwKr6zuOkdRNDM4eDZuOVloR1E
 https://www.youtube.com/watch?v=KuHrlgJ1Z7g (from min 30)

http://onevm-177.lal.in2p3.fr:8080/

Codalab sample competition bundle



Past seminars

WINTER/SPRING 2013

 Date  Time  Speaker/
Moderator
 Paper(s) discussed  Slides (optional)  Notes/ recording
 Thursday, Jan 31, 2013  9 am PT Alain-Jacques Valleron, Pierre Bougnères  Searching the environmental causes of complex diseases View Download
Dropbox link

 Thursday, May 9, 2013  9 am PT  Isabelle Guyon  Cause-effect pairs challenge View Download pptx w. voice over Summary
Meeting notes
 Thursday, May 16, 2013  9 am PT  Dominik Janzing  Information Geometry method to detecting cause-effect relationships View Download  
 Thursday, May 23, 2013  9 am PT  Romain Reuillon  OpenMOLE: a workflow system for massively distributed executions  Abstract Slides  
 Wednesday, June 5, 2013  9 am PT  Joris Mooij  MML methods for detecting cause-effect relationships  Paper View slides Slides(pdf)  
 Thursday, June 13, 2013  9 am PT  Kun Zhang  Identifying cause-effect relationships with non-linear functional models. View Download   

FALL 2012

 Date  Time  Speaker/
Moderator
 Paper(s) discussed  Slides (optional)  Notes/ recording
 Thursday, Sept. 27, 2012  9 am PT  Jennie Si  Computing with neural spikes View Download  Summary
 Tuesday, Oct. 2, 2012  9 am PT      Florin Popescu  Model Selection and ChaLearn    
 Thursday, Oct. 11, 2012  9 am PT  Lambert Schomaker  Model Selection in large-scale problems View Download   
 Thursday, Oct. 18, 2012  9 am PT  Kun Zhang  Non-linear functional causal models View Download   
 Thursday, Oct 25, 2012  9 am PT Manuel Gomez Rodriguez   Structure and Dynamics of Information Pathways in On-line Media 
View Download   
View Download  Summary
 Thursday, Nov. 1, 2012  9 am PT Leon Bottou    [preprint] Couterfactual Reasoning and Learning Systems  View Download  
Tuesday, Nov. 20, 2012  9 am PT Demian Battaglia Dynamic information routing in neuronal networks of the brain View Download   Summary
 Thursday, Nov. 29, 2012  9 am PT Navid Hassanpour A Dynamic Threshold Model of Collective Action in Social Networks
Canceled 
   
 Thursday, Dec. 20, 2012  9 am PT Susan Athey   A Structural Model of Sponsored Search Advertising Auctions  View Download   summary


SPRING 2012
We are alternated discussions on 2 themes: Model selection (led by the group ModelSelect) -- yellow weeks -- and Causal systems (led by the group CausaSimul) -- white weeks.

 Date  Time  Speaker/
Moderator
 Paper(s) discussed  Slides (optional)  Notes/ recording
 Thursday, March 8, 2012  9 am PT

 Isabelle Guyon  Model Selection beyond the Bayesian/Frequentist divide and model selection toolkit proposal    Notes
 Thursday, March 15, 2012  8 am PT Mikael Henaff 1) Benchmarking regulatory network reconstruction with GRENDEL
2) GeneNetWeaver: in silico benchmark generation and performance profiling of network inference methods
3) Revealing strengths and weaknesses of methods for gene network inference
   Notes
 Thursday, March 22, 2012  9 am PT  Florin Popescu  Optimal or preferred data splits.
Model selection for probabilistic clustering using cross-validated likelihood
 Figure  Summary
 Thursday, March 29  9 am PT  Isabelle Guyon          Description of the Virtual Laboratory and GLOP. Discussion on interfacing large "realistic" simulators.  Screen shots  
 Thursday, April 5  9 am PT  Gavin Cawley  Model selection - the best place to look for performance gains (see chapters 3 and 13 of "hands-on book")  Slides  
 Thursday, April 12  9 am PT  Richard Kennaway  When causation does not imply correlation: robust violations of the faithfulness axiom  Slides  Abstract
 Thursday, April 19  9 am PT  Kristin Bennett  Multi-level optimization:
 chapter 15 of "hands-on book"
and model selection for primal svm
   Notes
 Thursday, April 26  9 am PT  Ioannis Tsamardinos  Towards Integrative Causal Analysis of Heterogeneous Datasets and Studies  Slides(ppt)
Slides (pdf)
 
 Thursday, May 3  9 am PT  Alexander Statnikov  Automodelers for machine learning and predictive analytics
Academia:

1. An overview of the GEMS system/automodeler for analysis of microarray gene expression data and biomarker discovery
2. A related system/automodeler FAST-AIMS for analysis of mass-spectrometry proteomics data
Industry:
1. An overview of key players in predictive analytics and automodeling 
2. Google Prediction API 
3. Oracle Predictive Analytics 
4. IBM SPSS Modeler 
 go to the paper page to download the paper PDFs  Summary
 Thursday, May 10  9 am PT  Frederick Eberhardt

1) Combining Experiments to Discover Linear Cyclic Models 

with Latent Variables

 (short paper)

2) Learning Linear Cyclic Causal Models with Latent Variables (theory details)

 Slides (pdf)  
 Thusday, May 17  9 am PT  Juha Reunanen  Read Publications II and III of Juha's thesis. View slides
Download slides [ppt]

 Summary
 Thursday, May 24  9 am PT  Florin Popescu          Causality in time series
[download CiML book, PDF]
Slides View
Download
 
 Thursday, May 31  9 am PT  Marc Boullé Data grid models for preparation and modeling in supervised learning. (see chapter 5 of "hand-on book").
Nonparametric Edge Density Estimation in Large Graphs. 
 Slides  Summary
 Thursday, June 7  9 am PT  Jonas Peters Identifiability of Restricted Structural Equation Models  View Download   Summary
 Thursday, June 14  9 am PT  Hugo Jair Escalante  Full model selection with heuristic search (see chapter 14 of "hand-on book").
Ensemble particle swarm model selection.
 Slides pptx
 

Misc. planned seminars

 Date  Time  Speaker/
Moderator
 Paper(s) discussed  Slides (optional)  Notes/ recording
 Thursday, xx, 2013  9 am PT Ioannis Tsamardinos  Incorporating prior knowledge in the learning of BN or MAG

 Thursday, xx, 2013  9 am PT  Gavin Cawley Ryan Tibshirani and Robert Tibshirani. A Bias Correction for the Minimum Error Rate in Cross-Validation. Annals of Applied Statistics, Vol. 3, No. 1, 822-829, 2009. 
Full paper: View Download   
Thursday, xx, 2013  9 am PT  Thomas S Richardson 

 Single World Intervention Graphs (SWIGs): A Unification of the Counterfactual and Graphical Approaches to Causality

View Download   
 Thursday, xx, 2013  9 am PT  David Jensen and Marc Maier Learning causal models from relational data.
 
Thursday, xx, 2013  9 am PT  Olav Stetter, Demian Battaglia

Model-free reconstruction of excitatory neuronal connectivity from calcium imaging signals. PLoS Comput Biol 8, e1002653.


 
   
Núria Macià, Ester Bernadó-Mansilla, Albert Orriols-Puig, Tin Kam HoPattern Recognition 46 (3), 1054-1066

Learner excellence biased by data set selection: A case for data characterisation and artificial data set

  View Download
 Thursday May 12, 2016 8 am PT    Baiyu Chen  Overcoming Calibration Problems in Pattern Labeling with Pairwise Ratings