BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Learning polynomials with Neural Networks  - Aldo Pacchiano (Berke
 ley)
DTSTART:20160526T133000Z
DTEND:20160526T150000Z
UID:TALK66393@talks.cam.ac.uk
CONTACT:Yingzhen Li
DESCRIPTION:We study the effectiveness of learning low degree polynomials 
 using neural networks by the gradient descent method. While neural network
 s have been shown to have great expressive power\, and gradient descent ha
 s been widely used in prac- tice for learning neural networks\, few theore
 tical guarantees are known for such methods. In particular\, it is well kn
 own that gradient descent can get stuck at local minima\, even for simple 
 classes of target functions. In this paper\, we present several positive t
 heoretical results to support the effectiveness of neural networks. We foc
 us on two- layer neural networks where the bottom layer is a set of non-li
 near hidden nodes\, and the top layer node is a linear function\, similar 
 to Bar- ron (1993). We show that for a randomly initialized neural network
  with sufficiently many hidden units\, the generic gradient descent algori
 thm learns any low degree polynomial\, assuming we initialize the weights 
 randomly 
LOCATION:Engineering Department\, CBL Room 438
END:VEVENT
END:VCALENDAR