BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
X-WR-CALNAME:I2S Speaker Series: Dr. Xintao Wu
X-WR-TIMEZONE:Central Time (US & Canada)
BEGIN:VEVENT
DTSTAMP:20260607T080129Z
UID:tag:localist.com\,2008:EventInstance_45454635265802
DTSTART:20240212T190000Z
DTEND:20240212T210000Z
DESCRIPTION:Title: Achieving causal fairness in bandit based recommendation
  \n\nAbstract \n\nIn online recommendation\, customers arrive in a sequent
 ial and stochastic manner from an underlying distribution and the online d
 ecision model recommends a chosen item for each arriving individual condit
 ional on the state of the environment. It is imperative to develop online 
 recommendation algorithms to maximize the expected reward while achieving 
 user-side fairness for customers\, i.e.\, customers who share similar prof
 iles will receive a similar reward regardless of their sensitive attribute
 s and items being recommended. We study how to leverage offline data\, inc
 orporate causal inference\, and adopt soft intervention to model the item 
 selection strategy in contextual bandits. We present the d-separation base
 d UCB algorithm (D-UCB) which can reduce the amount of exploration needed 
 to achieve low cumulative regret\, and the fair causal bandit (F-UCB) for 
 achieving the counterfactual individual fairness. As the offline data ofte
 n contain confounding and selection biases\, ignoring these biases in caus
 al bandits could negatively affect the performance of online recommendatio
 n. We present approaches of estimating conditional causal effects and deri
 ving their bounds in the presence of compound biases. We further study how
  the derived causal bounds affect regret analysis in contextual bandits.  
  \n\nBiography \n\nDr. Xintao Wu is a professor and the Charles D. Morgan/
 Acxiom Endowed Graduate Research Chair and leads Social Awareness and Inte
 lligent Learning (SAIL) Lab in the Electrical Engineering and Computer Sci
 ence Department at the University of Arkansas.  He was a faculty member in
  College of Computing and Informatics at the University of North Carolina 
 at Charlotte from 2001 to 2014. He got his BS degree in Information Scienc
 e from the University of Science and Technology of China in 1994\, ME degr
 ee in Computer Engineering from the Chinese Academy of Space Technology in
  1997\, and Ph.D. in Information Technology from George Mason University i
 n 2001. Dr. Wu's major research interests include data mining\, privacy an
 d security\, fair machine learning\, and recently causal bandits and causa
 l representation learning. Dr. Wu has published over 180 scholarly paper
 s and received several paper awards including PAKDD'13 Best Application Pa
 per Award\, BIBM'13 Best Paper Award\, CNS'19 Best Paper Award\, and PAKDD
 '19 Most Influential Paper Award. He has also served on editorial boards o
 f several international journals and many conference program committees of
  data mining and AI.
GEO:38.952187;-95.264135
LOCATION:Nichols Hall\, Apollo Auditorium\, Rm 129
SUMMARY:I2S Speaker Series: Dr. Xintao Wu
URL;VALUE=URI:https://calendar.ku.edu/event/i2s_speaker_series_dr_xintao_wu
CATEGORIES:Lectures & Presentations
CATEGORIES:Academic
END:VEVENT
END:VCALENDAR