AI Paper Atlas

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

Simultaneous Confidence Bounds for Aggregated Effects via Exact Subset Optimization

Weihang Xu, Huajie Qian, Wotao Yin, Xinshang Wang

ICML 2026 · Learning theory and algorithms › Combinatorial and search optimization · approach: Predict-then-optimize learning

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 SubsetMaxT bound instead of BonferroniBonferroni · Synthetic dense homogeneous, m=50, k=50mean lower confidence bound: -1.62e+00 -> 3.10e-01Improved
use SubsetMaxT bound instead of ScheffeScheffe · Synthetic dense heterogeneous, m=100, k=100mean lower confidence bound: -8.08e-01 -> 7.31e-01Improved
use SubsetMaxT bound instead of ScheffeScheffe · Synthetic sparse homogeneous, m=50, k=5mean lower confidence bound: 2.18e+00 -> 2.25e+00Improved
use SubsetMaxT bound instead of BonferroniBonferroni · Hillstrom subgroup A/B, m=32, k=32mean lower confidence bound: 3.72e-02 -> 7.88e-02Improved
use SubsetMaxT bound instead of ScheffeScheffe · Criteo subgroup A/B, m=49, k=49mean lower confidence bound: 1.36e+00 -> 1.40e+00Improved
use SubsetMaxT bound instead of BonferroniBonferroni · OpenML-CC18 pipelines, m=50, k=50mean lower confidence bound: 1.91e-02 -> 3.79e-02Improved
impose the cardinality limit k=5 instead of unconstrained SubsetMaxT (ablation)SubsetMaxT (unconstr.) · Synthetic dense homogeneous, m=50, k=5mean lower confidence bound: -4.80e-01 -> -2.15e-01Improved
impose the cardinality limit k=5 instead of unconstrained SubsetMaxT (FPTAS) (ablation)SubsetMaxT (unconstr.) · Synthetic dense homogeneous, m=100, k=5mean lower confidence bound: -8.68e-01 -> -2.59e-01Improved
use unconstrained SubsetMaxT instead of Scheffe at large m small kScheffe · Synthetic dense homogeneous, m=100, k=5mean lower confidence bound: -3.06e-01 -> -8.68e-01Hurt
use FPTAS (eps=0.01) instead of exact MILPSubsetMaxT · Synthetic dense homogeneous, m=50, k=5mean lower confidence bound: -2.15e-01 -> -2.28e-01Hurt
use FPTAS (eps=0.01) instead of exact MILPSubsetMaxT · Synthetic sparse homogeneous, m=50, k=5mean lower confidence bound: 2.25e+00 -> 2.23e+00No clear effect
use SubsetMaxT instead of Bonferroni when k=1Bonferroni · ML pipelines simple injection, delta=0.02, k=1mean lower confidence bound: -0.005 -> -0.007Hurt

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