Cerfacs Enter the world of high performance ...

From 11 October 2021 to 14 October 2021

Machine learning for data science

nasri |  

Complete

Deadline for registration: 15 days before the starting date of each training
Duration : 4 days / (28 hours)

Before signing up, you may wish to report us any particular constraints (schedules, health, unavailability…)

at the following e-mail address : training@cerfacs.fr

Satisfaction index

In May 2019, 100% of participants were satisfied or very satisfied

(results collected from 17 respondents out of 18 participants, a response rate of 94%)

Abstract

This training course enables the participants to reinforce their theoretical and practical knowledge in order to implement machine learning techniques for the automatic analysis of data. The main statistical methods for data analysis are presented, both for data exploration (non-supervised learning) and for prediction (supervised learning). Each method is first presented and commented on a theoretical level, and then illustrated on numerical experiments run with public datasets using R and/or python/scikit-learn software.

Objective of the training

To know the main algorithms of automatic data analysis, and to know how to use them with R and/or python/scikit-learn.

Learning outcomes

The participants should be able to :

  • recognize the type of problem that they are facing (supervised or non-supervised learning, sequential learning, reinforcement learning…);
  • choose the right algorithm to use;
  • use an R on python implementation of this algorithm.

Teaching methods

The training is an alternation of theoretical presentations and practical work. A multiple choice question allows the final evaluation. The training room is equipped with computers, the work can be done in sub-groups of two people.

Target participants

This training session is for students, engineers, and computer scientists who wish to reinforce or extend their theoretical background and practical knowledge on automatic data analysis by statistical learning algorithms.

Prerequisites and registration

  • Basic knowledge in statistics: elementary probability, statistical tests, Gaussian linear model.
  • Basic knowledge in algorithmic and programming.
  • Install Python 2.7 with Anaconda, R 3.4.2 and IRkernel. Internet access during the sessions in order to get possible updates or to load additional libraries.
  • The training can take place in French or English depending on the audience, level B2 o f CEFR is required.

In order to verify that the prerequisites are satisfied, the following questionnaire must be completed. You need to get at least 75% of correct answers in order to be authorized to follow this training session. If you don’t succeed it, your subscription will not be validated. You only have two chances to complete it.

Questionnaire 1 https://goo.gl/forms/xL86TzPDFOC5r7ln1

After completing the pre-requisite tests and obtaining at least 75% correct answers, you can register:

Pre-registration

Referent teacher: Sébastien GERCHINOVITZ

Fee

  • Traines/PhDs/PostDocs : 280 € excl. tax
  • Cerfacs shareholders/CNRS/INRIA : 800 € excl. tax
  • Public : 1600 € excl. tax

Program

Every day from 9h to 17h30.

Morning: lecture; afternoon: hands-on sessions.

Day 1

General presentation of statistical machine learning and its main approachs. Comparison with traditional statistics and machine learning.
Unsupervised learning:
– Principal component analysis
– Agglomerative Hierarchical Clustering
– k-means, k-medoids and variants
– overview of other methods : Affinity Propagation, dbscan, etc.
Day 2
Supervised learning 1 / 2 :
– k nearest neighbors
– Gaussian linear model, logistic regression, model selection
– LASSO et variants
– Support Vector Machines
Day 3
Supervised learning 2 / 2 :
– Decision Trees
– Bagging, Random Forests, Boosting
– Neural networks, deep learning
Day 4
Sequential learning, multi-armed bandit problems
Super-learning and expert aggregation
Reinforcement learning (introduction)

Evaluation of learning

A final exam will be conducted during the training.

NEWS

Sophie Valcke from Cerfacs co-authored a new book on atmosphere-ocean modelling

CERFACS |  18 August 2021

new book "Atmosphere-Ocean Modelling - Couling and Couplers” by Prof. Carlos R Mechoso, Prof. Soon-Il An and Dr Sophie Valcke has just been published by World Scientific. The present book fills a void in the current literature by presenting a basic and yet rigorous treatment of how the models of the atmosphere and the ocean are put together into a coupled system. Details are available at  Abstract: Coupled atmosphere-ocean models are at the core of numerical climate models. There is an extraordinarily broad class of coupled atmosphere-ocean models ranging from sets of equations that can be solved analytically to highly detailed representations of Nature requiring the most advanced computers for execution. The models are applied to subjects including the conceptual understanding of Earth’s climate, predictions that support human activities in a variable climate, and projections aimed to prepare society for climate change. The present book fills a void in the current literature by presenting a basic and yet rigorous treatment of how the models of the atmosphere and the ocean are put together into a coupled system. The text of the book is divided into chapters organized according to complexity of the components that are coupled. Two full chapters are dedicated to current efforts on the development of generalist couplers and coupling methodologies all over the worldRead more


CERFACS to participate to the pitch competition “Falling walls lab”

CERFACS |  6 August 2021

Javier Crespo-Anadon, a PhD candidate from EU Marie-Curie project Annulight, has been selected to participate in the research pitch competition "Falling walls lab" (). This event will be held online on September 30th and it is organized in collaboration with the EU Marie Skłodowska-Curie Actions. Researchers from all over Europe compete to deliver the best research pitch presentation under the format "my thesis in 180 seconds".Read more

ALL NEWS