BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Trading-Off Payments and Accuracy in Online Classification  - Ciar
 a Pike-Burke (Imperial College London)
DTSTART:20231020T130000Z
DTEND:20231020T140000Z
UID:TALK206008@talks.cam.ac.uk
CONTACT:Qingyuan Zhao
DESCRIPTION:We consider online binary classification where in each round\,
  before making a prediction the learner can choose to ask some a number of
  stochastic experts for their advice. In contrast to the standard experts 
 problem\, we investigate the case where each expert needs to be paid befor
 e they provide their advice\, and that the amount we pay them directly inf
 luences the accuracy of their prediction through some unknown productivity
  function. In each round\, the learner must decide how much to pay each ex
 pert and then make a prediction. They incur a cost equal to a weighted sum
  of the prediction error and upfront payments for all experts. We introduc
 e an online learning algorithm and analyse its total cost compared to that
  of a predictor which knows the productivity of all experts in advance. In
  order to achieve this result\, we combine Lipschitz bandits and online cl
 assification with surrogate losses.\n\nJoint work with: Dirk van der Hoeve
 n\, Hao Qiu\, Nicolo Cesa-Bianchi
LOCATION:MR12\, Centre for Mathematical Sciences
END:VEVENT
END:VCALENDAR