AI Paper Atlas

Research map / Learning theory and algorithms / Combinatorial and search optimization

A Surrogate Objective Framework for Prediction+Programming with Soft Constraints

Kai Yan, Jie Yan, Chuan Luo, Liting Chen, Qingwei Lin, Dongmei Zhang

NeurIPS 2021 · Learning theory and algorithms › Combinatorial and search optimization · approach: Decision-focused optimization · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use the proposed surrogate soft-constraint objective instead of L1 two-stageprediction network · Synthetic LP, N=100, size (40, 40, 0)regret: 2.454 -> 2.258Improved
use the proposed objective instead of DFprediction network · Synthetic LP, N=1000, size (80, 80, 40)regret: 3.351 -> 2.781Improved
use SPO+ instead of two-stage L1/L2 lossesprediction network · Synthetic LP, N=100, size (80, 80, 40)regret: 4.786 -> 4.939Hurt
use DF instead of two-stage lossesprediction network · Synthetic LP, N=5000, size (80, 80, 40)regret: 2.224 -> 2.748Hurt
increase training set size from 100 to 5000all methods · Synthetic LP, size (80, 80, 0)regret: 5.200 -> 2.845Improved
increase training set size from 100 to 5000, gap to two-stageOurs vs L1 · Synthetic LP, size (40, 40, 0)regret: 2.454 -> 1.077Mixed
add 20 soft constraints to the LPOurs · Synthetic LP, N=1000, size 40 variablesregret: 1.346 -> 1.506Hurt
use the proposed objective instead of DFprediction network · Portfolio optimization, 50 equitiesregret (%): 4.016 -> 3.662Improved
use DF instead of L1/L2 two-stageprediction network · Portfolio optimization, 100 equitiesregret (%): 4.262 -> 3.500Improved
increase number of equities from 50 to 250Ours · Portfolio optimizationregret (%): 3.662 -> 3.212Improved
use the proposed objective instead of L1prediction network · Resource provisioning, alpha1/alpha2=100regret: 105.061 -> 20.829Improved
use the proposed objective instead of L1 when alpha1/alpha2=1prediction network · Resource provisioning, alpha1/alpha2=1regret: 4.267 -> 5.839Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy