English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The end-to-end predict-then-optimize framework, also known asdecision-focused learning, has gained popularity for its ability to integrateoptimization into the training procedure of machine learning models thatpredict the unknown cost (objective function) coefficients of optimizationproblems from contextual instance information. Naturally, most of the problemsof interest in this space can be cast as integer linear programs. In this work,we focus on binary linear programs (BLPs) and propose a new end-to-end trainingmethod for predict-then-optimize. Our method, Cone-aligned Vector Estimation(CaVE), aligns the predicted cost vectors with the cone corresponding to thetrue optimal solution of a training instance. When the predicted cost vectorlies inside the cone, the optimal solution to the linear relaxation of thebinary problem is optimal w.r.t. to the true cost vector. Not only does thisalignment produce decision-aware learning models, but it also dramaticallyreduces training time as it circumvents the need to solve BLPs to compute aloss function with its gradients. Experiments across multiple datasets showthat our method exhibits a favorable trade-off between training time andsolution quality, particularly with large-scale optimization problems such asvehicle routing, a hard BLP that has yet to benefit from predict-then-optimizemethods in the literature due to its difficulty.

CaVE: A Cone-Aligned Approach for Fast Predict-then-optimize with Binary Linear Programs