Know more

Our use of cookies

Cookies are a set of data stored on a user’s device when the user browses a web site. The data is in a file containing an ID number, the name of the server which deposited it and, in some cases, an expiry date. We use cookies to record information about your visit, language of preference, and other parameters on the site in order to optimise your next visit and make the site even more useful to you.

To improve your experience, we use cookies to store certain browsing information and provide secure navigation, and to collect statistics with a view to improve the site’s features. For a complete list of the cookies we use, download “Ghostery”, a free plug-in for browsers which can detect, and, in some cases, block cookies.

Ghostery is available here for free:

You can also visit the CNIL web site for instructions on how to configure your browser to manage cookie storage on your device.

In the case of third-party advertising cookies, you can also visit the following site:, offered by digital advertising professionals within the European Digital Advertising Alliance (EDAA). From the site, you can deny or accept the cookies used by advertising professionals who are members.

It is also possible to block certain third-party cookies directly via publishers:

Cookie type

Means of blocking

Analytical and performance cookies

Google Analytics

Targeted advertising cookies


The following types of cookies may be used on our websites:

Mandatory cookies

Functional cookies

Social media and advertising cookies

These cookies are needed to ensure the proper functioning of the site and cannot be disabled. They help ensure a secure connection and the basic availability of our website.

These cookies allow us to analyse site use in order to measure and optimise performance. They allow us to store your sign-in information and display the different components of our website in a more coherent way.

These cookies are used by advertising agencies such as Google and by social media sites such as LinkedIn and Facebook. Among other things, they allow pages to be shared on social media, the posting of comments, and the publication (on our site or elsewhere) of ads that reflect your centres of interest.

Our EZPublish content management system (CMS) uses CAS and PHP session cookies and the New Relic cookie for monitoring purposes (IP, response times).

These cookies are deleted at the end of the browsing session (when you log off or close your browser window)

Our EZPublish content management system (CMS) uses the XiTi cookie to measure traffic. Our service provider is AT Internet. This company stores data (IPs, date and time of access, length of the visit and pages viewed) for six months.

Our EZPublish content management system (CMS) does not use this type of cookie.

For more information about the cookies we use, contact INRA’s Data Protection Officer by email at or by post at:

24, chemin de Borde Rouge –Auzeville – CS52627
31326 Castanet Tolosan CEDEX - France

Dernière mise à jour : Mai 2018

Menu Logo Principal logo partenaire


StatMathAppli 2017

Statistics Mathematics and Applications

September 4 to 8, 2017
La Villa Clythia, Fréjus - Var (France)  

Aims :

The seminar "Statistics Mathematics and Applications" follows the seminar that held at Luminy each two years in 2006, 2004, ... and at Fréjus in 2008, 2010, 2011, 2013 and 2015.
The idea of this seminar is to give an opportunity to young statisticians from several countries to meet and present their work in an international meeting. Two series of invited lectures on mathematical statistics and their applications in the real life will be given by outstanding speakers.

Invited speakers :

This year, the invited lectures will be given by :

  • Francis BACH, INRIA Paris (
Francis Bach

Large-scale machine learning and convex optimization.
Many statistics, machine learning and signal processing problems are traditionally cast as convex optimization problems. A common difficulty in solving these problems is the size of the data, where there are many observations ("large n") and each of these is large ("large p"). In this setting, online algorithms such as stochastic gradient descent which pass over the data only once, are usually preferred over batch algorithms, which require multiple passes over the data. Given n observations/iterations, the optimal convergence rates of these algorithms are O(1/\sqrt{n}) for general convex functions and reaches O(1/n) for strongly-convex functions. 
In this tutorial, I will first present the classical results in stochastic approximation and relate them to classical optimization and statistics results. I will then show how the smoothness of loss functions may be used to design novel algorithms with improved behavior, both in theory and practice: in the ideal infinite-data setting, an efficient novel Newton-based stochastic approximation algorithm leads to a convergence rate of O(1/n) without strong convexity assumptions, while in the practical finite-data setting, an appropriate combination of batch and online algorithms leads to unexpected behaviors, such as a linear convergence rate for strongly convex problems, with an iteration cost similar to stochastic gradient descent.


Statistics with matrices, graphs, and tensors

Many modern datasets take the form of matrices or tensors. Applications range from
collaborative filtering to network analysis, to imaging. I will survey some basic statistical models, and
techniques to deal with them. Possible topics will include:
1) The spiked matrix model.
2) Semidefinite programming relaxations.
3) Z2 synchronization and the two-groups stochastic block model.
4) Sparse low-rank matrices, and the hidden clique problem.
5) Message passing algorithms
6) Estimating low-rank tensors.


All news
Symposium poster 2017

Symposium poster 2017

The 2017 edition is launched.
Read more


Read more

MIA-Jouy merged with MIG to create the MaIAGE Unit on January 1, 2015

A new Unit in INRA of Jouy en Josas
Read more

See all news