BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:IC Colloquium: A flexible framework for machine learning
DTSTART:20220331T100000
DTEND:20220331T110000
DTSTAMP:20260407T215712Z
UID:55659345b09865715b1a963d2456b28db5c1e3a95512dca6f59261e7
CATEGORIES:Conferences - Seminars
DESCRIPTION:By: Ferran Alet - MIT\nIC Faculty candidate\n\nAbstract\nIn th
 is last decade\, we have seen a lot of progress in AI and machine learning
  using a single recipe: given a task\, we train a single neural network to
  map inputs to outputs. In this talk\, I will show that this one-neural-ne
 twork-per-task framework can be extended to improve generalization. First\
 , I will describe modular meta-learning\, which achieves language-like gen
 eralization by training a set of composable neural modules. By having mult
 iple neural networks per task\, and multiple tasks per neural network\, we
  are able to reuse information and achieve bigger data and computational e
 fficiency. In the second part of my talk\, I will describe tailoring\, a g
 eneral way of encoding inductive biases in neural networks by optimizing u
 nsupervised objectives inside the prediction function\, essentially having
  one neural network per input. Finally\, I will describe my vision for cre
 ating a flexible ML framework that will enable training reinforcement lear
 ning policies within minutes rather than days\, solving complex search and
  discovery problems\, and improving our understanding of generalization in
  deep learning.\n\nBio\nFerran Alet is a PhD candidate at MIT CSAIL advise
 d by Leslie Kaelbling\, Tomas Lozano-Perez\, and Joshua Tenenbaum. His res
 earch is on machine learning and leverages techniques from meta-learning\,
  learning to search\, program synthesis\, and insights from mathematics an
 d the physical sciences. During his PhD\, he created the MIT Embodied Inte
 lligence Seminar\, mentored 17 students\, and won the MIT Outstanding Ment
 or award 2021. Ferran studied mathematics and physics in Barcelona thanks 
 to CFIS\, a program for doing two degrees\, where he was the valedictorian
  of his promotion. In college\, he participated in the ACM-ICPC programmin
 g contest\, being the most decorated in the history of his regional phase 
 (South Western Europe). In grad school\, he earned a “La Caixa” fellow
 ship and was responsible for the high-level planner of the MIT-Princeton t
 eam for the Amazon Robotics Challenge\, which won the stowing task in 2017
 .\n\nMore information
LOCATION:BC 420 https://plan.epfl.ch/?room==BC%20420 https://epfl.zoom.us/
 j/68824729051?pwd=aFVVSWNUM2l2UVBuSmtrSkNveUxVQT09
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR