August holidays

Youth and maker workshops

A first edition to learn game creation, electronics and connected objects.

Ages 8 to 13

Explorers

Create, play, understand

A playful journey to discover the digital world and create first creative projects.

Program

Discovering the digital world
Creating animations and mini-games
Introduction to technology and electronics
Final project presentation

Project examples

Mini-games
Interactive animations
Light-up objects
Creative challenges
Personalized projects

Ages 14 to 17

Makers

Imagine, build, experiment

A creative journey to explore programming, artificial intelligence and interactive technologies through concrete projects.

Program

Discovering generative artificial intelligence
Digital creation and creative programming
Introduction to technology and electronics
Final project presentation

Project examples

Creative mini-games
Interactive experiments
AI-assisted projects
Simple electronic objects
Personalized projects

Workshop dates

August 3 to 14, 2026

Registration is closed.

8 three-hour sessions. Places are limited to ensure good support.

Minimum: 6 participants per group

Maximum: 8 participants per group

Snack and water: provided by the laboratory

Computer: children may bring their own computer, but it is not required

Machine Learning Workshop

Machine Learning: understand neural networks by building them

Two three-hour sessionsSaturdays only

A practical workshop for students, self-taught learners, developers and artificial intelligence enthusiasts who want to understand the real foundations of neural networks and Deep Learning.

About this workshop

This workshop is not a TensorFlow or PyTorch course.

The goal is to understand the fundamental principles behind modern Machine Learning libraries.

Participants will progressively build the essential mechanisms of a neural network to understand how machine learning actually works.

Once these foundations are mastered, using libraries such as TensorFlow or PyTorch becomes much more natural.

What this workshop is

✅ An introduction to the foundations of Deep Learning
✅ A progressive understanding of neural networks
✅ An exploration of the main equations used in Machine Learning
✅ An approach focused on understanding rather than tool usage
✅ A step-by-step construction of the essential learning mechanisms

What this workshop is not

❌ A TensorFlow course
❌ A PyTorch course
❌ A simple use of existing libraries
❌ A series of tutorials to reproduce

Workshop specificity

The concepts studied are independent of the programming language used.

Participants may implement the exercises in the language of their choice, provided they know it well enough.

The examples and explanations remain applicable regardless of the language used.

Python
C#
Java
JavaScript / TypeScript
C++
Go
Rust
Any other language mastered by the participant

Session 1 Foundations of neural networks

Objective

Understand the fundamental principles of Machine Learning and progressively build a first neural network without using a specialized library.

Content

Introduction to Machine Learning

What is Machine Learning?
Types of problems and learning
Why do neural networks exist?

The artificial neuron

Biological inspiration
Inputs, weights and bias
Activation function
Producing a prediction

The perceptron

How it works
Limits of the simple perceptron
Introduction to multilayer networks

Understanding learning

Cost function
Concept of error
Why does a model learn?

Gradient descent

Intuition behind optimization
Slope of a function
Searching for a minimum

Backpropagation

General principle
Adjusting weights
Propagating error

Practice

Progressive construction of a simple neural network
Implementation without a Machine Learning library
Detailed analysis of every calculation step

Expected outcome

Explain how an artificial neuron works
Understand the main learning equations
Explain gradient descent and backpropagation
Build a simple neural network in their own programming language

Session 2 Foundations of convolutional neural networks (CNN)

Objective

Understand why convolutional networks revolutionized image processing and progressively build a first CNN.

Content

Why CNNs?

Limits of classic neural networks
Challenges related to images
The emergence of convolutional networks

Understanding a digital image

Pixels
Color channels
Matrix representation

Convolutions

Fundamental principle
Filters and kernels
Pattern detection
Feature extraction

Feature maps

Feature Maps
Progressive construction of representations

Pooling

Dimensionality reduction
Robustness to variations

CNN architecture

Layer stacking
Data flow through the network
Final classification

Practice

Building a simple CNN
Detailed analysis of calculations
Application to an image classification problem

Expected outcome

Explain how a convolutional network works
Understand the role of convolutions and pooling
Explain how a CNN extracts features from an image
Implement the fundamental principles of a CNN in their own programming language

Prerequisites

Know at least one programming language
Be comfortable with algebra and basic mathematical functions
Understand derivatives and the slope of a curve
Be willing to work with simple mathematical formulas
Bring a personal laptop. Laboratory computers are available in limited numbers

Registration validation

This workshop requires some programming and mathematics foundations.

A short placement test will be conducted before registration is confirmed to ensure that each participant can follow the training in good conditions.

Group size

Workshop held with at least 4 participants
Places limited to 6 participants

This intentionally small format allows personalized support and direct discussion with every participant.

Interested in registering?

Request a place