News : (reload for fresher news...) Lesson notes are now available in HTML format NB: this is the archive page of the course that took place last year! The new course is there.
Schedule : January-March 2019, on Monday mornings, 8h30 - 11h45 (3 hours with a 15 minute break) at CentraleSupelec
January 14th :Introduction: gap between practice and theory + Practical session on training GANs →install PyTorch, Jupyter and matplotlib → download the mini-project instructions →lesson summary
January 21st :Architectures + Guest talk by Therapixel →lesson summary →Presentation by Yaroslav Nikulin (Therapixel)
February 11th :Interpretability + Guest talk by Pierre Stock (Facebook FAIR Paris) →lesson summary →Presentation by Pierre Stock (Facebook FAIR, Paris)
February 18th :Deep Reinforcement Learning + Guest talk by Ludovic Denoyer (FAIR, Paris) →Deep RL slides
March 4th :Scaling Deep Learning + Guest talk by Rémi Cadène (LIP6) →lesson's slides →Presentation by Rémi Cadène (LIP 6)
March 11th :Small data, weak supervision and robustness / Modeling: exploiting known invariances, priors or physical properties + Practical session →3rd mini-project to be done individually and due for April 10th at 11 pm, to be submitted here (same password) →slides on small data →lesson summary on modeling
March 18th :Auto-ML / Auto-DeepLearning, by Isabelle Guyon's team + Discovery of the Auto-DL challenge →Presentation by Zhengying Liu (LRI, TAU team) →Instructions for the Auto-DL challenge
Location : amphi Janet (also named room E2.19) most often, in the Breguet building at CentraleSupelec. Bus stop: Moulon + 5 minute walk. Common transportation planification: vianavigo.
Mailing-list :
all announcements (last-minute changes, rooms, projects, etc.)
were made on a mailing-list no longer in use, check the webpage of the new course for this year's mailing-list.
Forum : a forum on piazza has been opened, meant for students to discuss and ask questions (password communicated on the mailing-list).