Nos « Tea Talks »

Nos conférences sont généralement présentées les vendredis à 10:30, salle AA-6214.

L’horaire de ces conférences et quelques unes des diapositives présentées sont disponibles (voir ci-dessous). Si vous trouvez un de ces articles intéressant, envoyez un courriel à .

Dates [M/D/Y]
Time
Speakers
Affiliation
Place
Titles
9/15/201711:00AMAnirudh GoyalMILAAA6124Z-Forcing
9/22/201712:00PMYoshua, Fred, MiriamMILAAA6214MILA new students meeting
9/29/201710:30AMIoannis MitliagkasMILAAA6124Understanding momentum dynamics for faster training, better scaling, and easier tuning
10/6/201710:30AMPhil BachmanMaluubaAA6124Metalearning and active learning -- like chocolate and peanut butter.
10/9/20172:00PMClement FarabetNVIDIAZ-209 Claire-McNicollDeep Learning Research at NVIDIA
10/13/201710:30AMHui JiangYork UniversityAA6124A New General Deep Learning Approach for Natural Language Processing
10/20/201710:30AMGuillaume AlainMILAAA6124Tutorial on the ML research workflow
10/27/2017----CANCELLED FOR ICLR DEADLINE
11/3/201710:30 AMPierre-Yves LablancheAIMS-South AfricaAA6214Machine Learning for Astrophysics
11/10/201710:30 AMRosemary Nan KeMILAAA6214Sparse Attentive Backtracking
11/17/201710:30 AMZac KentonMILAAA6214Three factors influencing minima in SGD
11/23/201710:30 AMMartin ArjovskyNYUAA6124A few geometrical insights into unsupervised learning.
11/24/201710:30 AMSergey PlisMRNAA6214Causality in complex time-series data
12/1/201710:30 AMRomain LarocheMicrosoft MaluubaAA6214Safe Policy Improvement with Baseline Bootstrapping
12/8/2017----CANCELLED FOR NIPS
12/15/2017
12/22/2017

Voir le Google doc

Diapositives disponibles