BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:America/New_York BEGIN:DAYLIGHT DTSTART:20250309T030000 TZOFFSETFROM:-0500 TZOFFSETTO:-0400 RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3 TZNAME:EDT END:DAYLIGHT BEGIN:STANDARD DTSTART:20251102T010000 TZOFFSETFROM:-0400 TZOFFSETTO:-0500 RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11 TZNAME:EST END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20250930T194208Z UID:3FC0012F-45A3-47D3-A064-E6A14EF397B9 DTSTART;TZID=America/New_York:20250930T110000 DTEND;TZID=America/New_York:20250930T123000 DESCRIPTION:Talk Title: Scientific Machine Learning for Electromagnetic Fie ld Computations\n\nAbstract\n\nA recent report by the US Department of Ene rgy defines the area of scientific machine learning as “a core component of artificial intelligence (AI) and a computational technology that can b e trained\, with scientific data\, to augment or automate human skills”\ , which has “the potential to transform science and energy research”. In this presentation\, we discuss the potential of scientific machine lear ning methods to problems in computational electromagnetics starting from s tandard electromagnetic structure analysis and multi-physics modeling in t he time-domain\, employing an unsupervised learning strategy based on Phys ics-Informed Neural Networks (PINN). PINNs directly integrate physical law s into their loss function\, so that the training process does not rely on the generation of training ground truth data from simulations (as in typi cal neural networks). We demonstrate the unconditionally stable solution o f coupled electromagnetic-thermal problems\, along with the modeling of fr equency selective surfaces and metasurfaces\, orders of magnitude faster t han the conventional finite-difference time-domain technique (FDTD)\, incl uding training time of the neural network.\n\nMoreover\, we demonstrate th e impact of machine learning on the computational modeling of radiowave pr opagation scenarios. We build convolutional neural network models that can process the geometry of indoor environments\, along with physics-inspired parameters\, to rapidly estimate received signal strength (RSS) maps. Emp hasis is placed on the generalizability of these models\, which is their a bility to "learn" the physics of radiowave propagation and produce accurat e modeling predictions in new geometries well beyond those included in the ir training set.\n\nSpeaker Bio\n\nProf. Costas Sarris is a Professor of E lectrical and Computer Engineering at the University of Toronto. His resea rch spans computational electromagnetics\, time-domain modeling\, wireless propagation models\, uncertainty quantification\, and scientific machine learning.\n\nHe is an IEEE Fellow and a Distinguished Lecturer of the IEEE Antennas and Propagation Society (2024–2026). His many honors include t he 2021 IET Premium Award for Best Paper in Microwaves\, Antennas & Propag ation and the 2013 IEEE MTT-S Outstanding Young Engineer Award. He has ser ved in numerous leadership roles\, including Editor-in-Chief of the IEEE J ournal on Multiscale and Multiphysics Computational Techniques (2019–202 4).\n\nDate & Time: Tuesday\, September 30\, 2025\, at 11:00 AM\nLocation: University of Waterloo\, EIT 3142\nSpeaker: Prof. Costas Sarris\, IEEE Fe llow\, University of Toronto\n\nOrganizers\n\n-\nIEEE KW MTT-S Student Bra nch Chapter\n\n-\nIEEE KW AP-S Student Branch Chapter\n\nCo-Organizers\n\n -\nIEEE KW Joint AP-S & MTT-S Chapter\n\n-\nIEEE KW Sensors Council Chapte r\n\n-\nIEEE KW Young Professionals\n\nEIT 3142\, 200 University Ave W\, O N\, Waterloo\, Ontario\, Canada\, N2L 3G1\, Virtual: https://events.vtools .ieee.org/m/501586 LOCATION:EIT 3142\, 200 University Ave W\, ON\, Waterloo\, Ontario\, Canada \, N2L 3G1\, Virtual: https://events.vtools.ieee.org/m/501586 ORGANIZER:amansoor@uwaterloo.ca SEQUENCE:55 SUMMARY:Distinguished Lecturer Talk by Distinguished Lecturer and IEEE Fell ow Professor Costas Sarris URL;VALUE=URI:https://events.vtools.ieee.org/m/501586 X-ALT-DESC:Description: <br /><h3 style="text-align: justify\;" data-start= "504" data-end="520">Talk Title: Scientific Machine Learning for Electroma gnetic Field Computations</h3>\n<h3 data-start="637" data-end="651">Abstra ct</h3>\n<p class="MsoNormal" style="margin-bottom: 12.0pt\; text-align: j ustify\;"><span lang="EN-US">A recent report by the US Department of Energ y defines the area of scientific machine learning as &ldquo\;a core compon ent of artificial intelligence (AI) and a computational technology that ca n be trained\, with scientific data\, to augment or automate human skills& rdquo\;\, which has &ldquo\;the potential to transform science and energy research&rdquo\;. In this presentation\, we discuss the potential of scien tific machine learning methods to problems in computational electromagneti cs starting from standard electromagnetic structure analysis and multi-phy sics modeling in the time-domain\, employing an unsupervised learning stra tegy based on Physics-Informed Neural Networks (PINN). PINNs directly inte grate physical laws into their loss function\, so that the training proces s does not rely on the generation of training ground truth data from simul ations (as in typical neural networks). We demonstrate the unconditionally stable solution of coupled electromagnetic-thermal problems\, along with the modeling of frequency selective surfaces and metasurfaces\, orders of magnitude faster than the conventional finite-difference time-domain techn ique (FDTD)\, <em>including training time of the neural network.</em> </sp an></p>\n<p class="MsoNormal" style="margin-bottom: 12.0pt\; text-align: j ustify\;"><span lang="EN-US">Moreover\, we demonstrate the impact of machi ne learning on the computational modeling of radiowave propagation scenari os. We build convolutional neural network models that can process the geom etry of indoor environments\, along with physics-inspired parameters\, to rapidly estimate received signal strength (RSS) maps. Emphasis is placed o n the <strong>generalizability</strong> of these models\, which is their a bility to "learn" the physics of radiowave propagation and produce accurat e modeling predictions in new geometries well beyond those included in the ir training set. </span></p>\n<p data-start="1404" data-end="1972">&nbsp\; </p>\n<h3 data-start="1505" data-end="1522">Speaker Bio</h3>\n<p data-star t="1523" data-end="1795"><strong data-start="1523" data-end="1546">Prof. C ostas Sarris</strong> is a Professor of Electrical and Computer Engineerin g at the University of Toronto. His research spans computational electroma gnetics\, time-domain modeling\, wireless propagation models\, uncertainty quantification\, and scientific machine learning.</p>\n<p data-start="179 7" data-end="2239">He is an <strong data-start="1806" data-end="1821">IEEE Fellow</strong> and a <strong data-start="1828" data-end="1911">Distingui shed Lecturer of the IEEE Antennas and Propagation Society (2024&ndash\;20 26)</strong>. His many honors include the 2021 IET Premium Award for Best Paper in <em data-start="1982" data-end="2018">Microwaves\, Antennas &amp\ ; Propagation</em> and the 2013 IEEE MTT-S Outstanding Young Engineer Awar d. He has served in numerous leadership roles\, including Editor-in-Chief of the <em data-start="2154" data-end="2224">IEEE Journal on Multiscale an d Multiphysics Computational Techniques</em> (2019&ndash\;2024).<br><br><s trong data-start="239" data-end="255">Date &amp\; Time:</strong> Tuesday\, September 30\, 2025\, at 11:00 AM<br data-start="295" data-end="298"><str ong data-start="298" data-end="311">Location:</strong> University of Water loo\, EIT 3142<br data-start="355" data-end="358"><strong data-start="358" data-end="370">Speaker:</strong> Prof. Costas Sarris\, IEEE Fellow\, Univ ersity of Toronto</p>\n<h3 data-start="226" data-end="246">&nbsp\;</h3>\n< h3 data-start="226" data-end="246">Organizers</h3>\n<ul data-start="247" d ata-end="327">\n<li data-start="247" data-end="287">\n<p data-start="249" data-end="287">IEEE KW MTT-S Student Branch Chapter</p>\n</li>\n<li data-s tart="288" data-end="327">\n<p data-start="290" data-end="327">IEEE KW AP- S Student Branch Chapter</p>\n</li>\n</ul>\n<h3 data-start="329" data-end= "352">Co-Organizers</h3>\n<ul data-start="353" data-end="459">\n<li data-s tart="353" data-end="391">\n<p data-start="355" data-end="391">IEEE KW Joi nt AP-S &amp\; MTT-S Chapter</p>\n</li>\n<li data-start="392" data-end="42 7">\n<p data-start="394" data-end="427">IEEE KW Sensors Council Chapter</p >\n</li>\n<li data-start="428" data-end="459">\n<p data-start="430" data-e nd="459">IEEE KW Young Professionals</p>\n</li>\n</ul>\n<p>&nbsp\;</p> END:VEVENT END:VCALENDAR