BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:DAYLIGHT
DTSTART:20260308T030000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:EDT
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20261101T010000
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:EST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260330T031831Z
UID:AA9AA109-846B-4E65-B0BB-5C2359F57037
DTSTART;TZID=America/New_York:20260502T110000
DTEND;TZID=America/New_York:20260502T121500
DESCRIPTION:This session explores the role of Explainable Artificial Intell
 igence (XAI) in improving transparency and trust in machine learning syste
 ms applied to complex real-world problems.\n\nThe first talk focuses on so
 ftware defect prediction\, addressing the growing complexity of modern sof
 tware systems where early defect detection is essential for maintaining re
 liability and reducing maintenance costs. The presented framework integrat
 es advanced data preprocessing\, class imbalance handling\, and the XGBoos
 t algorithm with explainability techniques\, enabling accurate defect pred
 iction while providing interpretable insights into the software metrics th
 at contribute most to defect formation.\n\nThe second talk examines anomal
 y detection in financial transactions\, a critical task for fraud preventi
 on and regulatory compliance. The proposed approach leverages attention-en
 hanced Variational Autoencoders (VAEs) combined with interpretability meth
 ods such as SHAP to identify anomalous transaction patterns. Beyond detect
 ing suspicious activity\, the framework provides transaction-level explana
 tions that help analysts understand the factors influencing anomaly classi
 fication.\n\nTogether\, these presentations demonstrate how explainable AI
  methods can bridge the gap between high-performing machine learning model
 s and the transparency required for real-world adoption in both software e
 ngineering and financial security.\n\nSpeaker(s): Srikanth\, Sowjanya\n\nV
 irtual: https://events.vtools.ieee.org/m/546377
LOCATION:Virtual: https://events.vtools.ieee.org/m/546377
ORGANIZER:aj.williams@ieee.org
SEQUENCE:50
SUMMARY:Explainable AI for Reliable Systems and Financial Security
URL;VALUE=URI:https://events.vtools.ieee.org/m/546377
X-ALT-DESC:Description: &lt;br /&gt;&lt;p data-start=&quot;222&quot; data-end=&quot;405&quot;&gt;This sessi
 on explores the role of &lt;strong data-start=&quot;256&quot; data-end=&quot;301&quot;&gt;Explainabl
 e Artificial Intelligence (XAI)&lt;/strong&gt; in improving transparency and tru
 st in machine learning systems applied to complex real-world problems.&lt;/p&gt;
 \n&lt;p data-start=&quot;407&quot; data-end=&quot;922&quot;&gt;The first talk focuses on &lt;strong dat
 a-start=&quot;433&quot; data-end=&quot;463&quot;&gt;software defect prediction&lt;/strong&gt;\, address
 ing the growing complexity of modern software systems where early defect d
 etection is essential for maintaining reliability and reducing maintenance
  costs. The presented framework integrates advanced data preprocessing\, c
 lass imbalance handling\, and the &lt;strong data-start=&quot;726&quot; data-end=&quot;737&quot;&gt;
 XGBoost&lt;/strong&gt; algorithm with explainability techniques\, enabling accur
 ate defect prediction while providing interpretable insights into the soft
 ware metrics that contribute most to defect formation.&lt;/p&gt;\n&lt;p data-start=
 &quot;924&quot; data-end=&quot;1418&quot;&gt;The second talk examines &lt;strong data-start=&quot;949&quot; da
 ta-end=&quot;996&quot;&gt;anomaly detection in financial transactions&lt;/strong&gt;\, a crit
 ical task for fraud prevention and regulatory compliance. The proposed app
 roach leverages &lt;strong data-start=&quot;1094&quot; data-end=&quot;1148&quot;&gt;attention-enhanc
 ed Variational Autoencoders (VAEs)&lt;/strong&gt; combined with interpretability
  methods such as &lt;strong data-start=&quot;1196&quot; data-end=&quot;1204&quot;&gt;SHAP&lt;/strong&gt; t
 o identify anomalous transaction patterns. Beyond detecting suspicious act
 ivity\, the framework provides transaction-level explanations that help an
 alysts understand the factors influencing anomaly classification.&lt;/p&gt;\n&lt;p 
 data-start=&quot;1420&quot; data-end=&quot;1666&quot;&gt;Together\, these presentations demonstra
 te how &lt;strong data-start=&quot;1466&quot; data-end=&quot;1613&quot;&gt;explainable AI methods ca
 n bridge the gap between high-performing machine learning models and the t
 ransparency required for real-world adoption&lt;/strong&gt; in both software eng
 ineering and financial security.&lt;/p&gt;
END:VEVENT
END:VCALENDAR