BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:Mexico/General
BEGIN:STANDARD
DTSTART:20221030T010000
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20251028T034933Z
UID:63B6FA76-F410-49DD-B410-F75D7811E62D
DTSTART;TZID=Mexico/General:20251022T130000
DTEND;TZID=Mexico/General:20251022T140000
DESCRIPTION:Generative Adversarial Networks (GANs) implement Machine Learni
 ng (ML) algorithms that have the ability to address competitive resource\n
 allocation problems together with detection and mitigation of anomalous be
 havior. In this talk\, we discuss their use in next-generation NextG) comm
 unications within the context of cognitive networks to address:\n\ni) spec
 trum sharing\,\nii) detecting anomalies\, and\niii) mitigating security at
 tacks.\n\nGANs have the following advantages. First\, they can learn and s
 ynthesize field data\, which can be costly\, time consuming\, and nonrepea
 table. Second\, they enable pre-training classifiers by using semisupervis
 ed data. Third\, they facilitate increased resolution. Fourth\, they enabl
 e recovering corrupted bits in the spectrum. The talk will provide basics 
 of GANs\, a comparative discussion on different kinds of GANs\, performanc
 e measures for GANs in computer vision and image processing as well as wir
 eless applications\, a number of datasets for wireless applications\, perf
 ormance measures for general classifiers\, a survey of the literature on G
 ANs for i)–iii) above\, some simulation results\, and future resear
 ch directions. In the spectrum sharing problem\, connections to cognitive 
 wireless networks are established. Simulation results show that a particul
 ar GAN implementation is better than a convolutional autoencoder for an ou
 tlier detection problem in spectrum sensing.\n\nCo-sponsored by: Instituto
  Politécnico Nacional - Unidad Profesional Interdisciplinaria de Ingenier
 ía "Alejo Peralta"\n\nSpeaker(s): Ender Ayanoglu\n\nBldg: Unidad Profesio
 nal Interdisciplinaria de Ingeniería "Alejo Peralta"\, Instituto Politéc
 nico Nacional\, Calle 11 Sur 12122\, San Francisco Mayorazgo\, Puebla\, Pu
 ebla\, Mexico\, 72480
LOCATION:Bldg: Unidad Profesional Interdisciplinaria de Ingeniería "Alejo 
 Peralta"\, Instituto Politécnico Nacional\, Calle 11 Sur 12122\, San Fran
 cisco Mayorazgo\, Puebla\, Puebla\, Mexico\, 72480
ORGANIZER:jvazquezbu@hotmail.com
SEQUENCE:61
SUMMARY:Machine Learning in NextG Networks via Generative Adversarial Netwo
 rks
URL;VALUE=URI:https://events.vtools.ieee.org/m/508427
X-ALT-DESC:Description: <br /><p class="MsoNormal"><span style="color: rgb(
 19\, 19\, 19)\; font-family: Verdana\, sans-serif\; font-size: 11pt\;">Gen
 erative Adversarial Networks (GANs) implement Machine Learning (ML)&nbsp\;
 algorithms that have the ability to address competitive resource<br>alloca
 tion problems together with detection and mitigation of&nbsp\;anomalous be
 havior. In this talk\, we discuss their use in&nbsp\;next-generation&nbsp\
 ; NextG) communications within the context of cognitive networks to addres
 s:</span></p>\n<p class="MsoNormal"><span style="color: rgb(19\, 19\, 19)\
 ; font-family: Verdana\, sans-serif\; font-size: 11pt\;">i)&nbsp\;spectrum
  sharing\,<br>ii) detecting anomalies\, and<br>iii) mitigating security at
 tacks.</span></p>\n<p class="MsoNormal"><span style="color: rgb(19\, 19\, 
 19)\; font-family: Verdana\, sans-serif\; font-size: 11pt\;">GANs have the
  following&nbsp\;advantages. First\, they can learn and synthesize field d
 ata\, which can&nbsp\;be costly\, time consuming\, and nonrepeatable. Seco
 nd\, they enable&nbsp\;pre-training classifiers by using semisupervised da
 ta. Third\, they&nbsp\;facilitate increased resolution. Fourth\, they enab
 le recovering&nbsp\;corrupted bits in the spectrum. The talk will provide 
 basics of GANs\,&nbsp\;a comparative discussion on different kinds of GANs
 \, performance&nbsp\;measures for GANs in computer vision and image proces
 sing as well as&nbsp\;wireless applications\, a number of datasets for wir
 eless applications\,&nbsp\;performance measures for general classifiers\, 
 a survey of the&nbsp\;literature on GANs for i)&acirc\;&euro\;&ldquo\;iii)
  above\, some simulation results\, and&nbsp\;future research directions. I
 n the spectrum sharing problem\,&nbsp\;connections to cognitive wireless n
 etworks are established. Simulation&nbsp\;results show that a particular G
 AN implementation is better than a&nbsp\;convolutional autoencoder for an 
 outlier detection problem in spectrum&nbsp\;sensing.</span></p>
END:VEVENT
END:VCALENDAR