BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20251016T135136Z
LOCATION:Grand Hall M/N
DTSTART;TZID=America/Chicago:20251016T171000
DTEND;TZID=America/Chicago:20251016T173000
UID:HFESAM_ASPIRE 2025_sess129_LECT347@linklings.com
SUMMARY:Machine Learning Considerations for Trust Predictions in HRI: Exam
 ination across Labels, Sex Differences, and Sensor Modalities
DESCRIPTION:Aakash Yadav and Ranjana Mehta (University of Wisconsin - Madi
 son)\n\nAdvances in AI and robotics increase human-robot interactions, but
  trust classification via neurophysiological data remains uncertain. Addit
 ionally, there is a limited understanding of how sex differences, model ch
 oice, data modality, and trust binarization influence trust prediction out
 comes. We collected data from thirty-eight participants who interacted wit
 h a UR10 robot to perform a planetary gear assembly task under reliable an
 d unreliable robot conditions. The dataset comprises neural activation, fu
 nctional connectivity, gaze entropy features, and behavioral measures such
  as trust ratings. Classical ML methods (AdaBoost, XGBoost, Naive Bayes Cl
 assifier, Logistic Regression, Random Forest, K-Nearest Neighbors, Multi-L
 ayer Perceptron, Support Vector Machine, and Decision Tree) were evaluated
  using 5-fold cross-validation with accuracy and F1 score as metrics. Our 
 results show 1) the prediction was better for male (acc = 0.935, mae = 0.0
 65) participants than female participants, 2) random forest performed bett
 er across all modalities and sexes (acc = 0.802, mae = 0.198), 3) midpoint
 -point binarization performed the best (acc = 0.898, mae = 0.102), and 4) 
 fNIRS-HbR data modality led to better prediction (acc = .898, mae = 102). 
 These insights underscore the need to explore trust development across div
 erse demographics and behaviors, informing effective HRI system design.\n\
 nTrack: Augmented Cognition\n\nSession Chairs: Jason Sanders (San José Sta
 te University) and Michael Hildebrandt (Institute for Energy Technology, H
 alden Project)\n\n
END:VEVENT
END:VCALENDAR