BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Practical Decisions in Neural Network Design - Joe Yan\, Magdalene
  College
DTSTART:20180214T190000Z
DTEND:20180214T193000Z
UID:TALK101368@talks.cam.ac.uk
CONTACT:Matthew Ireland
DESCRIPTION:The performance of machine learning algorithms is growing quic
 kly due to the development of hardware and training model and method. In s
 ome areas\, machine learning even has a better performance than natural hu
 mans. The neural network is a powerful model for machine learning.  To bui
 ld a practical neural network\, there are different choices that must be m
 ade at different stages. For deeper and more complex architectures\, these
  minor differences could leads to an enormous model converge speed differe
 nce. This talk is a discussion about the general neural network. The conte
 nt includes different choices and trade-offs on loss function\, activation
  function\, parameter update method and forward/backward propagation by bo
 th maths model and practical training evidence. 
LOCATION:Wolfson Hall\, Churchill College
END:VEVENT
END:VCALENDAR