BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:DAYLIGHT
DTSTART:20230312T030000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:EDT
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20231105T010000
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:EST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20230929T155550Z
UID:E852EAF1-A903-43BA-BF7D-7667B41092D0
DTSTART;TZID=America/New_York:20230926T140000
DTEND;TZID=America/New_York:20230926T150000
DESCRIPTION:Abstract: With the superior capability of discovering intricate
  structure of large data sets\, machine learning has been widely applied i
 n various areas including wireless networking. Modern wireless networks ha
 ve seen a tremendous increase in size and complexity\, where the capacity 
 optimization urgently calls for an innovative and efficient computing para
 digm. This talk shares our recent studies on how to exploit deep learning 
 for significant performance gain in wireless network optimization. We star
 t with a discussion about our vision of the supremacy of machine learning 
 (ML)\, trying to fundamentally understand where and how the ML based appro
 aches show advantages versus the conventional modeling-based approaches. W
 e then demonstrate that ML has the unique capability to identify latent st
 ructure information\, embedded in historical solved optimization instances
  but invisible to traditional algorithms\, leading to new algorithms that 
 can greatly mitigate the computation overhead while maintain a good perfor
 mance. New ML techniques have been developed to extract and leverage the s
 tructure information from three perspectives: topology-level\, algorithm-l
 evel\, and application-level.\n\nCo-sponsored by: Weihua Zhuang\n\nSpeaker
 (s): Yu Cheng\, \n\nRoom: 4152\, Bldg: EIT\, 200 Univ. Ave. West\, Univers
 ity of Waterloo\, Waterloo\, Ontario\, Canada\, N2L 3G1
LOCATION:Room: 4152\, Bldg: EIT\, 200 Univ. Ave. West\, University of Water
 loo\, Waterloo\, Ontario\, Canada\, N2L 3G1
ORGANIZER:wzhuang@uwaterloo.ca
SEQUENCE:6
SUMMARY:Seminar: To Find the Supremacy of Machine Learning in Wireless Netw
 ork Optimization
URL;VALUE=URI:https://events.vtools.ieee.org/m/372466
X-ALT-DESC:Description: <br /><p><strong>Abstract</strong>: With the superi
 or capability of discovering intricate structure of large data sets\, mach
 ine learning has been widely applied in various areas including wireless n
 etworking. Modern wireless networks have seen a tremendous increase in siz
 e and complexity\, where the capacity optimization urgently calls for an i
 nnovative and efficient computing paradigm. This talk shares our recent st
 udies on how to exploit deep learning for significant performance gain in 
 wireless network optimization. We start with a discussion about our vision
  of the supremacy of machine learning (ML)\, trying to fundamentally under
 stand where and how the ML based approaches show advantages versus the con
 ventional modeling-based approaches. We then demonstrate that ML has the u
 nique capability to identify latent structure information\, embedded in hi
 storical solved optimization instances but invisible to traditional algori
 thms\, leading to new algorithms that can greatly mitigate the computation
  overhead while maintain a good performance. New ML techniques have been d
 eveloped to extract and leverage the structure information from three pers
 pectives: topology-level\, algorithm-level\, and application-level.&nbsp\;
 </p>
END:VEVENT
END:VCALENDAR