Conference item icon

Conference item

Stable matching with evolving preferences

Abstract:

We consider the problem of stable matching with dynamic preference lists. At each time-step, the preference list of some player may change by swapping random adjacent members. The goal of a central agency (algorithm) is to maintain an approximately stable matching, in terms of number of blocking pairs, at all time-steps. The changes in the preference lists are not reported to the algorithm, but must instead be probed explicitly. We design an algorithm that in expectation and with high probability maintains a matching that has at most O((log n)2) blocking pairs.

Publication status:
Published
Peer review status:
Peer reviewed

Actions

Access Document

Files:
Publisher copy:
10.4230/LIPIcs.APPROX-RANDOM.2016.36

Authors

More by this author
Institution:
University of Oxford
Division:
MPLS
Department:
Computer Science
Role:
Author


Publisher:
Schloss Dagstuhl
Host title:
Approximation, Randomization, and Combinatorial Optimization. Algorithms and Techniques (APPROX/RANDOM 2016)
Journal:
LIPICS More from this journal
Volume:
60
Pages:
36:1--36:13
Series:
Leibniz International Proceedings in Informatics (LIPIcs)
Publication date:
2016-09-01
Acceptance date:
2016-06-22
DOI:
ISSN:
1868-8969
ISBN:
9783959770187


Keywords:
Pubs id:
pubs:632973
UUID:
uuid:b3927b91-0d43-414f-8395-1f4df689c469
Local pid:
pubs:632973
Source identifiers:
632973
Deposit date:
2016-07-10
ARK identifier:

Terms of use


Views and Downloads






If you are the owner of this record, you can report an update to it here: Report update to this record

TO TOP