Machine learning - 5AMCS529

A+Augmenter la taille du texteA-Réduire la taille du texteImprimer le documentEnvoyer cette page par mail

Number of hours
- Lectures : 15.0
- Tutorials : -
- Laboratory works : 6.0
- Projects : -
- Internship : -
- Written tests : -
ECTS : 3.5
Officials : Laurent LEFEVRE

Goals

Investigate the main models of Artificial Intelligence. The courses and practical work allow students to discover and create AI models by themselves, starting from elementary components (such as the formal neuron) towards the most complex models (such as deep neural networks). Unlike black-box approach, the course aims at understanding in detail how the models studied work.

Content

The history of AI and its various models
neural networks: from formal neuron to deep learning
unsupervised networks and self-organizing maps
reinforcement learning
robotics and developmental learning

Prerequisites

A preliminary experience (basic knowledge) in programming (Java and Python)

Tests

The evaluation includes two labs (50%) and a final exam (50%).
TP (lab): two sessions with evaluation of in-session work and written report
E1: session 1 individual exam (practical and theoretical)
E2: session 2 individual exam (practical and theoretical)
DM: (in case of confinment) individual homework with report and online oral exam

The exam is given in english only