Solving Hard Stable Matching Problems via Local Search and Cooperative Parallelization

Danny Munera (1) , Daniel Diaz (1) , Salvador Abreu (2) , Francesca Rossi (3) , Vijay Saraswat (4) , Philippe Codognet (5, 6)

1 CRI - Centre de Recherche en Informatique de Paris 1
2 Universidade de Évora
3 Unipd - Università degli Studi di Padova = University of Padua
4 IBM Thomas J. Watson Research Center
5 JFLI - Japanese French Laboratory for Informatics
6 UPMC - Université Pierre et Marie Curie - Paris 6

Stable matching problems have several practical applications. If preference lists are truncated and contain ties, finding a stable matching with maximal size is computationally difficult. We address this problem using a local search technique, based on Adaptive Search and present experimental evidence that this approach is much more efficient than state-of-the-art exact and approximate methods. Moreover, parallel versions (particularly versions with communication) improve performance so much that very large and hard instances can be solved quickly.

Domains: Distributed, Parallel, and Cluster Computing [cs.DC]

Danny Munera, Daniel Diaz, Salvador Abreu, Francesca Rossi, Vijay Saraswat, et al.. Solving Hard Stable Matching Problems via Local Search and Cooperative Parallelization. 29th AAAI Conference on Artificial Intelligence, Jan 2015, Austin, TX, United States. ⟨hal-01144214⟩