News : (reload for fresher news...)
- This webpage is about last year's course. Check the one for 2021.
- Comments welcomed! about this class here
General information about the course: presentation slides
The previous course (January-March 2019) can be found there; the new course will follow more or less the same lines, though not exactly.
Teaching team :
Most lectures: Guillaume Charpiat
Practical sessions: Victor Berger & Wenzhuo Liu
Schedule : January-March 2020, most often on Monday mornings, 9h - 12h15 (3 hours with a 15 minute break) at CentraleSupelec, amphi Janet (also named room E2.19, in the Breguet building), except when indicated:
January 13th : Deep learning vs. classical machine learning and optimization + Practical session: hyperparameters and training basics →Bring your laptop! →Environment used in practical sessions → Practical session: topic and notebooks for Exercise 1, Exercise 2, Exercise 3 and Exercise 4 with its dataset; submit full report (for all exercises together) here (password communicated on mailing-list). →Practical tips and tricks to train neural networks →Lesson summary
February 17th : in amphi Ampère (D1.06, same building) : Small or noisy data: forms of weak supervision → Practical session on active learning: jupyter notebook and submission site →Lesson summary
February 24th : Deep Reinforcement Learning, guest talk by Olivier Teytaud (Facebook FAIR Paris), + guest talk by Yaroslav Nikulin (Therapixel) → Slides by Olivier Teytaud: Evolutionary computer vision, Nevergrad, Polygames → Yaroslav Nikulin's slides
February 27th[Thursday!], 13h45 - 17h : Modeling: deep learning and physics (exploiting known invariances, priors or physical properties), by Michele Alessandro Bucci (LRI, TAU team) and Lionel Mathelin (LIMSI, Paris-Sud) →Practical session on deep learning for physics (due by March 12th); submission site → Lionel Mathelin's slides available on request →Last year's course on the topic
March 9th : Variational approach + Auto-ML / Auto-DeepLearning, guest talk by Zhengying Liu (LRI, TAU team, Isabelle Guyon's group) → Zhengying Liu's talk → Practical session on generative models; submission site
March 16th : Guarantees? Neural Tangent Kernel, generalization and formal proofs
Note that the schedule is irregular.
Location : Bus stop: Moulon + 7 minute walk to the Breguet building. Common transportation planification: vianavigo.
Mailing-list :
all announcements (last-minute changes, rooms, projects, etc.) will be made on the following mailing-list; to subscribe, replace /info/ with /subscribe/ in the following URL (yes, this is a cheap antispam) :
https://sympa.inria.fr/sympa/info/deep_learning_in_practice_2020 . Make sure that
you do receive an confirmation email; if not, try subscribing with another email address, or contact me.
Auditors (auditeurs libres) are welcome; just subscribe to the mailing-list as well.
Forum : a forum on piazza has been opened, meant for students to discuss and ask questions (password communicated on the mailing-list).