BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260611T063041EDT-0168UlpfXb@132.216.98.100
DTSTAMP:20260611T103041Z
DESCRIPTION:Title: An Adaptive Algorithm to Multi-armed Bandit Problem with
  High-dimensional Covariates\n\n\n	Abstract:\n\n\nThis work studies an impo
 rtant sequential decision making problem known as the multi-armed bandit p
 roblem with covariates. Under a linear bandit framework with high-dimensio
 nal covariates\, we propose a general arm allocation algorithm that integr
 ates both arm elimination and randomized assignment strategies. By employi
 ng a class of high-dimensional regression methods for coefficient estimati
 on\, the proposed algorithm is shown to have near optimal finite-time regr
 et performance under a new study scope that requires neither a margin cond
 ition nor a reward gap condition for competitive arms. Based on synergisti
 cally verified benefit of the margin\, our algorithm exhibits an adaptive 
 performance that automatically adapts to the margin and gap conditions\, a
 nd attains the optimal regret rates under both study scopes\, without or w
 ith the margin\, up to a logarithmic factor. The proposed algorithm also s
 imultaneously generates useful coefficient estimation output for competiti
 ve arms and is shown to achieve both estimation consistency and variable s
 election consistency. Promising empirical performance is demonstrated thro
 ugh two real data evaluation examples in drug dose assignment and news art
 icle recommendation.\n\n\n	Speaker\n\n\nDr. Wei Qian is an Assistant Profes
 sor of Statistics at the University of Delaware. His research interests in
 clude high-dimensional statistics\, sequential decision making\, actuarial
  statistics\, online recommendation\, and data science applications. I am 
 particularly interested in investigating statistical machine learning meth
 ods for analysis of complex and big data stemming from applied problems.\n
 \nZoom Link\n\nMeeting ID: 843 0865 5572\n\nPasscode: 690084\n
DTSTART:20210205T203000Z
DTEND:20210205T213000Z
SUMMARY:Dr. Wei Qian (University of Delaware)
URL:https://www.mcgill.ca/mathstat/channels/event/dr-wei-qian-university-de
 laware-328157
END:VEVENT
END:VCALENDAR