BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:Machine learning in Physics\, Chemistry and Materi
 als discussion group (MLDG)
SUMMARY:Graph Convolutional Networks for Atomic Structures
  - Rachel Kurchin\, Carnegie Mellon University
DTSTART;TZID=Europe/London:20201207T170000
DTEND;TZID=Europe/London:20201207T173000
UID:TALK154633AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/154633
DESCRIPTION:I am developing two packages in the Julia programm
 ing language to facilitate graph-based machine lea
 rning in atomic systems: crystals\, surfaces\, mol
 ecules\, etc. In this talk\, I will first give a b
 rief tutorial on the math behind graph convolution
 \, then introduce the packages: ChemistryFeaturiza
 tion.jl for building and featurizing the atomic gr
 aphs\, and AtomicGraphNets.jl for building and tra
 ining the models. I will also compare the capabili
 ties and performance of my code to the Python-base
 d implementation of a similar model.
LOCATION:virtual ZOOM meeting ID: 263 591 6003\, Passcode: 
 000042\, https://us02web.zoom.us/j/2635916003?pwd=
 ZlBEQnRENGwxNmJGMENGMWxjak5nUT09
CONTACT:Bingqing Cheng 
END:VEVENT
END:VCALENDAR