BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Energy\, entropy and the physics of deep learning - Dr Alpha Lee\,
  Cavendish Laboratory
DTSTART:20190130T200000Z
DTEND:20190130T210000Z
UID:TALK116815@talks.cam.ac.uk
CONTACT:Akshat Pandey
DESCRIPTION:Deep learning has achieved beyond-human accuracy in a plethora
  of challenging tasks\, ranging from image recognition to gameplay. Noneth
 eless\, why deep learning "works” is thus far an open question. In my ta
 lk\, I will argue that physical concepts such as energy and entropy allow 
 us to explain the surprising efficacy of deep learning. I will show that s
 tochastic gradient descent\, a machine learning algorithm that is commonly
  used\, can be mapped to the familiar physics of Brownian motion albeit wi
 th a spatially anisotropic noise that is crucial to its success. Moreover\
 , theoretical tools used to analyse spin glasses and energy landscapes giv
 e us important insights about the structure of loss functions that are typ
 ically encountered in deep learning.
LOCATION:Wolfson Lecture Theatre\,  Department of Chemistry\, Lensfield Ro
 ad
END:VEVENT
END:VCALENDAR