When are Kalman-Filter Restless Bandits Indexable?
Chris Dance, Tomi Silander
We study the restless bandit associated with an extremely simple scalar Kalman lter
model in discrete time. Under certain assumptions, we prove that the problem is
in the sense that the
is a non-decreasing function of the relevant belief state.
In spite of the long history of this problem, this appears to be the rst such proof. We use
, which are particular binary strings
intimately related to
NIPS, Montréal, Canada, December 7-12, 2015.