The distribution of the best arm identification task on the user's devices offers several advantages for application purposes: scalability, reduction of deployment costs and privacy. We propose a distributed version of the algorithm Successive Elimination using a simple architecture based on a single server which synchronizes each task executed on the user's devices. We show that this algorithm is near optimal both in terms of transmitted number of bits and in terms of number of pulls per player. Finally, we propose an extension of this approach to distribute the contextual bandit algorithm Bandit Forest, which is able to finely exploit the user's data while guaranteeing the privacy.
Help us improve arXiv so we can better serve you. Take our user survey (survey closes April 27, 8PM EDT).
from cs.AI updates on arXiv.org http://ift.tt/1XmBDce
via IFTTT
No comments:
Post a Comment