INFORMATION - Training & Research

"Artificial Intelligence" training given by O. Philip
(DAT Solutions)

INITIAL TRAINING (January 9 to 12, 2023) FULL

Target: Masters, PhD students, post-docs and permanent staff of the institute's member laboratories.

Organization: 4-day session (24h)

How to register :
Places are limited to 20 people FULL SESSION
Free and mandatory registration.
Requests will be taken into account as they come in.

Documents to send: document " IMAGING - Registration IA 2022-23 " by
by email to the following address: cecile.lavoute@univ-amu.fr

ADVANCED TRAINING (5 to 8 December 2022)

Target: PhD students, post-docs and permanent staff of the Institute's member laboratories
(pre-requisite: INITIAL training)

Organization: 4 days session (24h)

How to register :
Places are limited to 20 people
Free and mandatory registration.
Requests will be taken into account as they come in.

Requirements to register :
General knowledge required:

• Machine learning (supervised and unsupervised learning, generalization, overlearning, underlearning, linear and logistic regression, cost function, gradient descent);
• Artificial neural networks (dense layer, convolution layer, residual shortcut, activation function, back-propagation);
• Auto-encoder, U-Net, GAN;
• Programming basics (Python), basic knowledge of TensorFlow or PyTorch.

Concepts covered:

• Self-supervised learning;
• Diffusion model;
• Transformers.

Documents to be sent: document " IMAGING - Registration IA 2022-23 " by mail
by email to the following address: cecile.lavoute@univ-amu.fr