Program > Program

Scientific activities:

The Statlearn Spring School will welcome several lectures:

  • Aurélie BOISBUNON - Datascientist, MyDataModels, Nice (France)
    • Practical applications of the L1 penalty. The L1 penalty has arisen a lot of interest in the statistical and machine learning communities in the 90's and early 2000's, due to its ability to both select relevant variables and/or features and estimate their corresponding weights at the same time. Despite its decrease in popularity with the recent advances in deep learning, it can still be very useful in many real-life applications. We will first introduce the main concepts of the L1 penalty and its convex and nonconvex variants (e.g. Elastic net, MCP, SCAD, log penalty, ...) : sparsity, bias, regularization path. Then, we will discuss how it can be applied in practice in different contexts beyond linear regression/classification, ranging from random feature generation to signal and image processing.
  • Claire BOYER - Assistant professor in Statistics, Sorbonne Université, Paris (France)
    • Benign overfitting & double descent.
    • Inference & learning with missing values.
  • Julien MAIRAL - Researcher in machine learning, INRIA, Grenoble (France)
    • Large-scale optimisation. Continuous optimization has been central to machine learning for a long time. In these lectures, we are interested in continuous problems with a particular "large-scale" structure that prevents us from using generic optimization toolboxes or simply vanilla first- or second-order gradient descent methods. In such a context, all of these tools suffer indeed either from too large complexity per iteration, or too slow convergence, or both, which has motivated the machine learning community to develop dedicated algorithms. We will introduce several of such techniques, and in particular focus on stochastic optimization, which plays a crucial role for applying machine learning techniques to large datasets.
  • Brendan Murphy - Full professor in Statistics,  University College, School of Mathematics & Statistics, Dublin (Ireland)
    • Model-based statistical learning with networks

Poster sessions, with some local food, are planned during the week where participants are encouraged to present their own work. They are a great opportunity for young researchers to present their work.

Social activities:

Local food and drinks will be proposed during the welcome drink on Monday, the poster sessions and the workshop dinner on Thursday.

On Wednesday, we will go to the Calanques de Piana and visit this wonderful UNESCO World Heritage Site! It will not be a hike, but bringing good walking shoes is highly recommanded.




Photos: P-A Mattei

Online user: 1 RSS Feed