BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:Asia/Kolkata
BEGIN:STANDARD
DTSTART:19451014T230000
TZOFFSETFROM:+0630
TZOFFSETTO:+0530
TZNAME:IST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260910T030414Z
UID:541C12EB-53D5-4D78-8197-3C0F1F279061
DTSTART;TZID=Asia/Kolkata:20260923T090000
DTEND;TZID=Asia/Kolkata:20260923T110000
DESCRIPTION:Knowledge Inclusion in Machine Learning: Adaptive Physics-Infor
 med Neural Networks\nSpeaker: Prof Saman Halgamuge\, FIEEE\nPopular Large 
 Language Models are Machine Learning (ML) models first trained on a large 
 Copus of textual general context subsequently fine-tuned to be user friend
 ly task specific. The performance of all machine learning models including
  Deep Neural Networks (DNNs) depend on the context available to train them
 . This context may be consisting of two forms: specific context from sampl
 e-level experimental data\, and general context from knowledge accumulated
  and curated over the human history. Current DNNs rarely consider this gen
 eral context. In our experience\, neither form of context is sufficient al
 one: experimental data are sensitive to dataset specific noise and lack br
 oader contextual grounding\, while curated knowledge may lack the resoluti
 on required for the specific problem. Consequently\, most DNNs relying sol
 ely on experimental data risk learning spurious correlations rather than u
 nderlying context. Here we introduce Knowledge Inclusive Machine Learning 
 (KIML)\, a paradigm that integrates both context types within a unified an
 alytical pipeline. KIML combines experimental data with various types of g
 eneral context: existing mathematical models (equations)\, literature-deri
 ved representations from research paper data bases and structured knowledg
 e graphs etc.\n\nSpeaker(s): \, Saman\n\nSt Joseph’s University\, 36 Lal
 bagh Road\, Bangalore\, Karnataka\, India
LOCATION:St Joseph’s University\, 36 Lalbagh Road\, Bangalore\, Karnataka
 \, India
ORGANIZER:rohitverma.kgp@gmail.com
SEQUENCE:8
SUMMARY:IEEE DV Talk by Prof Saman Halgamuge
URL;VALUE=URI:https://events.vtools.ieee.org/m/576835
X-ALT-DESC:Description: <br /><div><strong>Knowledge Inclusion in Machine L
 earning: Adaptive Physics-Informed Neural Networks</strong></div>\n<div><s
 trong>Speaker: Prof Saman Halgamuge\, FIEEE</strong></div>\n<div>Popular L
 arge Language Models are Machine Learning (ML) models first trained on a l
 arge Copus of textual general context subsequently fine-tuned to be user f
 riendly task specific. The performance of all machine learning models incl
 uding Deep Neural Networks (DNNs) depend on the context available to train
  them. This context may be consisting of two forms: specific context from 
 sample-level experimental data\, and general context from knowledge accumu
 lated and curated over the human history. Current DNNs rarely consider thi
 s general context.&nbsp\; In our experience\, neither form of context is s
 ufficient alone: experimental data are sensitive to dataset specific noise
  and lack broader contextual grounding\, while curated knowledge may lack 
 the resolution required for the specific problem. Consequently\, most DNNs
  relying solely on experimental data risk learning spurious correlations r
 ather than underlying context. Here we introduce Knowledge Inclusive Machi
 ne Learning (KIML)\, a paradigm that integrates both context types within 
 a unified analytical pipeline. KIML combines experimental data with variou
 s types of general context: existing mathematical models (equations)\, lit
 erature-derived representations from research paper data bases and structu
 red knowledge graphs etc.</div>
END:VEVENT
END:VCALENDAR