Mathematical Methods of Modern Statistics
July 10 – 14, 2017

Harald Cramér’s  Mathematical Methods of Statistics is a landmark both for Mathematics and Statistics.  2016  was the 70th anniversary of its first publication. Again, after 70 years, mathematics and statistics seem to be falling more and more apart. Numerous statistical procedures are based on ad hoc methods supported by intensive computer-assisted simulations, while mathematicians and mathematical statisticians not always know, face  and confront the real issues of modern statistics.  A much better understanding between statisticians and mathematicians  is essential for the development of both these fields: a new Mathematical Methods of Modern Statistics remains to be written. 

Our aim is to gather in Luminy world class statisticians who use excellent mathematics, and mathematicians who work in their area of interest. We plan to discuss recent achievements in modern statistics, requiring an extensive use of deep mathematical methods and models. We would like to put emphasis on detailed exposition of mathematics  behind the tools used in statistics.

The major unifying topic will be the analysis of large dimensional  data.
The conference will include the following topics:

1. methods of multiple testing      
2. model selection theory                 
3. notions of model sparsity           
4. regularization techniques             
5. missing data treatments             
6. hierarchical and graphical models
7. modern non parametric Bayes methods (BNP)
8. interactions between the above topics
9. random matrices
10. mathematical methods applied in the above topics

The subject of the conference will be viewed from:

(a) the frequentist perspective
(b) the Bayesian perspective

Felix Abramovich (Tel Aviv University)   From model selection in GLM to sparse logistic classification
Sylvain Arlot (Université Paris-Sud)   Analysis of some purely random forests
Yannick Baraud (Université de Nice-Sophia-Antipolis)   How to make Bayes estimators robust
Jean-Marc Bardet (Université Paris 1)   Statistical analysis of causal ane processes
Yoav Benjamini (Tel Aviv University)   A survey of challenges in large scale selective inference
Philippe Biane (Université Paris Est)   Free probability and random matrices
Lucien Birgé   (Université Pierre-et-Marie-Curie)   How to make Bayes estimators robust
Małgorzata Bogdan (Wrocław University)    Sorted L-One Penalized Estimation
Bryc Włodzimierz (University of Cincinnati)   Cauchy-Stieltjes families with polynomial variance functions
Emmanuel Candès (Stanford University)   A new read of the knockos framework : new statistical tools for replicable selections
Ismael Castillo (Université Pierre-et-Marie-Curie)   Uniform estimation of a class of random graph functionals
Mathias Drton (University of Washington, Seattle)   Regularized score matching for graphical models : Non-Gaussianity and missing data
David Dunson  (Duke University)   Bayesian manifold learning
Christophe Giraud (Université Paris Sud)   Optimal clustering with convex optimisation
Ruth Heller (Tel Aviv University)   Inference following selection by aggregate level hypothesis testing
Hideyuki Ishi (Nagoya University)   Wishart laws for a wide class of regular convex cones
Julie Josse (Ecole Polytechnique)   Multiple imputation for continuous and categorical variables with low-rank methods
Rafał Latała (Warsaw University)   Comparison of weak and strong moments for vectors with independent coordinates
Steffen Lauritzen (University of Copenhagen)   Total positivity and conditional independence
Oleg Lepski (Université d’Aix-Marseille)   Estimation in the convolution structure density model
Gérard Letac (Université de Toulouse)   A generalisation of the Sabot-Tarrès integral and the multivariate normal law with non positive correlations
Clément Marteau (Université Claude Bernard, Lyon)   Parameter recovery in two-component contamination mixtures​
Hélène Massam (York University, Toronto)   Existence of the maximum likelihood estimate in discrete graphical models
Nicolai Meinshausen (ETH Zürich)   Causal Dantzig : fast inference in linear structural equation models with hidden variables under additive interventions
​Takaaki Nomura ( Kyushu University, Fukuoka)   Homogeneous open convex cones : recent results
Yann Ollivier ((Université Paris-Sud)   Real-time gradient descents for learning dynamical systems
Dominique Picard (Université Paris 7)   Clustering high dimensional data with sparsity
Etienne Roquain (Université Pierre-et-Marie-Curie)   Post hoc inference via joint family-wise error rate control
Judith Rousseau (Université Paris Dauphine)   Bayesian nonparametric inference for multivariate Hawkes processes
Chiara Sabatti (Stanford University)    Selective inference in genetics
Richard Sameworth (Cambridge University)   Efficient multivariate entropy estimation via k-nearest neighbour distances
David Siegmund (Stanford University)  Detection and estimation of local signals
Jonathan Taylor (Stanford University)   Inferactive data analysis
Surya Tokdar (Duke University)   Linear quantile regression : joint estimation of quantile planes
Sara Van de Geer (ETH Zürich)   Estimating equations and sharp oracle results
Aad Van der Vaart  (Leiden University)   Statistical estimation of a network model
Jean Phillippe Vert (ENS Ulm, Paris)   Learning on the symmetric group
Nicolas Verzelen (INRA Montpellier)   On optimal graphon estimation
Jacek Wesołowski (Warsaw University)   Morality and immorality for discrete graphical models
Daniel Yekutieli (Tel Aviv University)    Confidence Intervals for the CDF from « noiy » iid samples
​​