We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Mathematics > Optimization and Control

Title: Relaxed Indexability and Index Policy for Partially Observable Restless Bandits

Authors: Keqin Liu
Abstract: This paper addresses an important class of restless multi-armed bandit (RMAB) problems that finds a broad application area in operations research, stochastic optimization, and reinforcement learning. There are $N$ independent Markov processes that may be operated, observed and offer rewards. Due to the resource constraint, we can only choose a subset of $M~(M<N)$ processes to operate and accrue reward determined by the states of selected processes. We formulate the problem as a partially observable RMAB with an infinite state space and design an algorithm that achieves a near-optimal performance with low complexity. Our algorithm is based on a generalization of Whittle's original idea of indexability. Referred to as the relaxed indexability, the extended definition leads to the efficient online verifications and computations of the approximate Whittle index under the proposed algorithmic framework.
Subjects: Optimization and Control (math.OC)
Cite as: arXiv:2107.11939 [math.OC]
  (or arXiv:2107.11939v3 [math.OC] for this version)

Submission history

From: Keqin Liu [view email]
[v1] Mon, 26 Jul 2021 03:22:14 GMT (1908kb)
[v2] Fri, 30 Jul 2021 10:21:19 GMT (1908kb)
[v3] Wed, 22 Feb 2023 07:51:33 GMT (2324kb)

Link back to: arXiv, form interface, contact.