Contrastive Losses and Solution Caching for Predict-and-Optimize

Maxime Mulamba, Jayanta Mandi, Michelangelo Diligenti, Michele Lombardi, Victor Bucarey Lopez, Tias Guns

Onderzoeksoutput: Conference paperResearch

6 Citaten (Scopus)

Samenvatting

Many decision-making processes involve solving a combinatorial optimization problem with uncertain input that can be estimated from historic data. Recently, problems in this class have been successfully addressed via end-to-end learning approaches, which rely on solving one optimization problem for each training instance at every epoch. In this context, we provide two distinct contributions. First, we use a Noise Contrastive approach to motivate a family of surrogate loss functions, based on viewing non-optimal solutions as negative examples. Second, we address a major bottleneck of all predict-and-optimize approaches, i.e. the need to frequently recompute optimal solutions at training time. This is done via a solver-agnostic solution caching scheme, and by replacing optimization calls with a lookup in the solution cache. The method is formally based on an inner approximation of the feasible space and, combined with a cache lookup strategy, provides a controllable trade-off between training time and accuracy of the loss approximation. We empirically show that even a very slow growth rate is enough to match the quality of state-of-the-art methods, at a fraction of the computational cost.

Originele taal-2English
TitelContrastive Losses and Solution Caching for Predict-and-Optimize
RedacteurenZhi-Hua Zhou
UitgeverijInternational Joint Conferences on Artificial Intelligence
Pagina's2833-2840
Aantal pagina's8
ISBN van elektronische versie9780999241196
DOI's
StatusPublished - 19 aug 2021
Evenement30th International Joint Conference on Artificial Intelligence (IJCAI-21): IJCAI-21 - Canada, Montreal, Canada
Duur: 21 aug 202126 aug 2021
https://ijcai-21.org/

Publicatie series

NaamIJCAI International Joint Conference on Artificial Intelligence
ISSN van geprinte versie1045-0823

Conference

Conference30th International Joint Conference on Artificial Intelligence (IJCAI-21)
Verkorte titelIJCAI 2021
Land/RegioCanada
StadMontreal
Periode21/08/2126/08/21
Internet adres

Vingerafdruk

Duik in de onderzoeksthema's van 'Contrastive Losses and Solution Caching for Predict-and-Optimize'. Samen vormen ze een unieke vingerafdruk.

Citeer dit