Descriptif

This course will provide an introduction to deep learning,from the perceptorn to the recent advances in deep architectures.The topics of the course will include MLPs, convolutional and recurrent architectures, autoencoders, etc. and will rely on Python and Keras for the practical part (exercises). The course will attempt at providing a balanced theoretical-practical approach for all the above topics.Prerequisites are knowledge in probability for the theoretical part and knowledge of Python (and possibly of Jupyter tebooks) for the practical part.