Number of hours
- Lectures 8.0
- Projects 0
- Tutorials 0
- Internship 0
- Laboratory works 16.0
- Written tests 0
ECTS
ECTS 2.0
Goal(s)
This course is an introduction to artificial intelligence: it presents the broad outlines of supervised learning, whether regression or classification. The aim is to present basic methods and algorithms. By the end of the course, students will be able to set up a simple AI chain dedicated to regression or classification, and deploy the model on a raspberry pi-type target device.
Contact Dawood AL CHANTI, Bertrand RIVETContent(s)
This course presents: linear regression, SVM classification, simple neural network architectures (perceptron, multi-layer perceptron, convolutional networks, auto-encoders and variational auto-encoders). The importance of training data is also addressed through data bias, unbalanced classes, cross-validation, etc.
Prerequisites
Scientific computing and digital signal processing.
Notion of optimization, digital convolution and digital correlation.
Programming in Python.
Test
SESSION NORMALE:
Types d'évaluation : CC + QCM
Evaluation non-rattrapable:
Contrôle continu : rapports de BE à déposer sur chamilo aux dates indiquées et en respectant les consignes imposées
Evaluation non-rattrapable:
QCM : à faire en suivant les consignes imposées et les dates imposées
Documents non autorisés
SESSION DE RATTRAPAGE:
examen individuel écrit de 1h30
Conditions : "Tous les documents sont interdits.
L’usage des téléphones portable, des appareils de communication et des calculatrices est interdit."
N1 = 50% CC1 + 50% QCM1
N2 = 50% CC1 + 50% DS2