BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.2.3.1//EN
TZID:Asia/Kolkata
X-WR-TIMEZONE:Asia/Kolkata
BEGIN:VEVENT
UID:81@cds.iisc.ac.in
DTSTART;TZID=Asia/Kolkata:20241118T093000
DTEND;TZID=Asia/Kolkata:20241118T103000
DTSTAMP:20241114T144103Z
URL:https://cds.iisc.ac.in/events/mtech-research-thesis-defense-hybrid-cds
 -18-november-2024-learning-multiple-initial-conditions-using-physics-infor
 med-neural-networks-pinns/
SUMMARY:Mtech Research Thesis Defense: HYBRID: CDS: 18\, November 2024 "Lea
 rning Multiple Initial Conditions Using Physics-Informed Neural Networks (
 PINNs)"
DESCRIPTION:DEPARTMENT OF COMPUTATIONAL AND DATA SCIENCES\nMtech Research T
 hesis Defense\n\n\n\nSpeaker : Mr. Mahesh Tom\nS.R. Number : 06-18-01-10-2
 2-21-1-20317\nTitle : "Learning Multiple Initial Conditions Using Physics-
 Informed Neural Networks (PINNs)"\nThesis examiner : Prof. Nagaiah Chamaku
 ri\nResearch Supervisor: Prof. Sashikumaar Ganesan\nDate &amp\; Time : Nov
 ember 18\, 2024 (Monday) at 9:30 AM\nVenue : The Thesis Défense will be h
 eld on HYBRID Mode\n\n# 102 CDS Seminar Hall /MICROSOFT TEAMS\n\nPlease cl
 ick on the following link to join the Thesis Defense: MS Teams link\n\n\n\
 nABSTRACT\nPhysics-Informed Neural Networks (PINNs) and their variants hav
 e emerged as tools for solving differential equations in the past few year
 s. Although several variants of PINNs have been proposed for time-dependen
 t partial differential equations (PDEs)\, the majority of these physics-in
 formed approaches are based on solving a problem for a single set of initi
 al conditions. In this work\, we consider one-dimensional time-dependent P
 DEs and focus on solving multiple initial conditions (ICs) with a single n
 etwork simultaneously. Trying to solve multiple ICs in a single network pr
 esents certain challenges\, such as spectral bias\, that we address in our
  work. We also look at how our approach performs in the FastVPINNs framewo
 rk to solve multiple ICs using Variational Physics-Informed Neural Network
 s (VPINNs). The choice of activation functions is crucial in the performan
 ce of a network\; hence we also test the influence of various activation f
 unctions on FastVPINNs for some standard test cases. While training multip
 le ICs\, we also look at the impact of the network parameters and how they
  contribute to each trained task via an ablation study.\n\nOnce we have a 
 fully trained model that works on multiple ICs\, incorporating new ICs wit
 hout having to retrain all the previous ICs is a challenging task\, and a 
 brute-force way of training all the ICs again is not always feasible. To t
 his end\, we explore the usage of elastic weight consolidation (EWC)\, a r
 egularization technique that is used in continual learning\, and study its
  effect on PINNs for training new ICs.\n\n\n\nALL ARE WELCOME
CATEGORIES:Events,Thesis Defense
END:VEVENT
BEGIN:VTIMEZONE
TZID:Asia/Kolkata
X-LIC-LOCATION:Asia/Kolkata
BEGIN:STANDARD
DTSTART:20231119T093000
TZOFFSETFROM:+0530
TZOFFSETTO:+0530
TZNAME:IST
END:STANDARD
END:VTIMEZONE
END:VCALENDAR