Artificial Intelligence

CentraleSupélec’s Artificial Intelligence Summer School, now at its second edition, includes lectures, tutorials and practicals enabling students to broaden their knowledge in fields such as Machine Learning, Deep and Reinforcement Learning, Computer Vision and Learning in Networks. Throughout the programme students will participate in a team-based competition on the topics addressed by the school.

Skills in Python and programming are required for the programme and if you are not familiar with Python, you will need to learn the basics prior to the start of the Summer School. In order to prove their level in Python, students will be asked to provide in the registration section a project previously completed demonstrating their programming skills.

For more information regarding tuition fees & accomodation, please visit our Practical Information Page 

Programme description

Summer School programme description

Introduction to Artificial Intelligence (Céline Hudelot)

This course will give an overview of the large field of Artificial Intelligence by first tracing its history and then by presenting the main Artificial Intelligence approaches: reflex-based, state-based, variable-based and logic-based. Students will apply these approaches during a practical lab session.

Introduction to Machine Learning (Richard Combes)

This course presents the fundamentals of statistical learning theory, from fundamental limits, to algorithms, their analysis and practical implementation. Important concepts such as Vapnik-Chervonenkis theory, Empirical Risk Minimization, Stochastic Gradient Descent algorithms and Kernel methods will be presented in details. A lab session will enable students to apply the theoretical contents to data.

Introduction to Deep Learning (Pablo Piantanida)

This course will provide students with the principles of representation learning and deep learning by covering the following subjects: Neural Networks, Backpropagation and stochastic gradient optimisation, Auto-encoders, Hyper-parameters and training tricks for neural networks, regularization, Deep Belief Networks and Deep Boltzmann Machines. Students will apply these approaches during a practical lab session.

Introduction to Reinforcement learning (Edouard Oyallon)

Reinforcement learning algorithms permit to teach an agent to optimize how to get a reward. This can be helpful for solving games such as backgammon, but more recently it was applied to Atari games or the game of Go and has led to impressive results. This is thanks to a combination with deep learning techniques. During this class, students will review elementary properties of reinforcement learning. A lab session will be provided as well.

Machine Learning on Networks (Fragkiskos Malliaros)

Networks (or graphs) have become ubiquitous as data from diverse disciplines can naturally be mapped to graph structures. Characteristic examples include social networks (e.g., Facebook, Twitter), information networks (e.g., the Web) as well as technological networks (e.g., the Internet). The problem of extracting meaningful information from large scale graph data in an efficient and effective way has become crucial and challenging with several important applications and towards this end, graph mining and learning methods constitute prominent tools. The goal of this course is to present recent and state-of-the-art methods and algorithms for analyzing, mining and learning large-scale network data, as well as their practical applications in various domains (e.g., the web, social networks, recommender systems).

Introduction to Computer Vision (Hugues Talbot and Maria Vakalopoulou)

Networks (or graphs) have become ubiquitous as data from diverse disciplines can naturally be mapped to graph structures. Characteristic examples include social networks (e.g., Facebook, Twitter), information networks (e.g., the Web) as well as technological networks (e.g., the Internet). The problem of extracting meaningful information from large scale graph data in an efficient and effective way has become crucial and challenging with several important applications and towards this end, graph mining and learning methods constitute prominent tools. The goal of this course is to present recent and state-of-the-art methods and algorithms for analyzing, mining and learning large-scale network data, as well as their practical applications in various domains (e.g., the web, social networks, recommender systems).

Artificial Intelligence System Architecture & building block (Anna Shillabeer)

The lecture will look into the notion of intelligence and compare this to what a computer does by using examples of expert systems especially in health and cybersecurity. Secondly, it will identify the components (e.g. machine learning, neural nets, analytics) of an ‘intelligent’ system (the high level architecture) and clarify what is the function of each. Thirdly, it will look into the strengths and weaknesses of intelligent systems in mission critical applications. Finally, it will try to assess what are the ethical and legal considerations of AI applications.

Constrained Deep Networks for Weakly Supervised Learning: Models and Optimization (Ismail Ben Ayed)

Recently, weakly supervised learning has drawn tremendous research interests in computer vision. The purpose is to mitigate the lack of full and laborious annotations in dense prediction tasks, e.g., semantic segmentation. In this talk, I will discuss some recent developments in this direction, focusing on how to enforce various types of constraints and priors on convolutional neural networks (CNNs), which can leverage unlabeled data, guiding training with domain-specific knowledge. I will discuss several key technical aspects in the context of CNNs with partial labels, including constrained optimization, conditional random fields and computational tractability. I will emphasize how more attention should be paid to optimization methods, going beyond standard gradient descent. The talk includes various illustrations and applications, which show how constrained CNNs can approach full-supervision performances while using fractions of the full ground-truth labels.

Please note that the schedule is subject to change

Academic Staff

Academic Staff

Fragkiskos D. Malliaros

Fragkiskos D. Malliaros

Assistant Professor & Organizer

Ismail Ben Ayed

Ismail Ben Ayed

Associate Professor

Richard Combes

Richard Combes

Assistant Professor

Céline Hudelot

Céline Hudelot

Professor

Edouard Oyallon

Edouard Oyallon

Assistant Professor

Pablo Piantanida

Pablo Piantanida

Associate Professor

Anna Shillabeer

Anna Shillabeer

Professor

Hugues Talbot

Hugues Talbot

Professor

Maria Vakalopoulou

Maria Vakalopoulou

Assistant Professor

TBD

TBD

TBD

Register Today!

Fill in the online registration form