AI Paper Atlas

Research map / Architectures and efficiency / Neural architecture search

Learning to Rank Learning Curves

Martin Wistuba, Tejaswini Pedapati

ICML 2020 · Architectures and efficiency › Neural architecture search · approach: AutoML pipeline search · 6 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 LCRankNet for early termination instead of training every configuration to completionRandom NAS in the NASNet search space · CIFAR-100Time (GPU hours): 1021 -> 11Improved
Use LCRankNet for early termination instead of training every configuration to completionRandom NAS in the NASNet search space · CIFAR-100Regret: 0.00 -> 0.00No clear effect
Use LCRankNet for early termination instead of training every configuration to completionRandom NAS in the NASNet search space · CIFAR-10Time (GPU hours): 1023 -> 20Improved
Use LCRankNet for early termination instead of training every configuration to completionRandom NAS in the NASNet search space · Fashion-MNISTRegret: 0.00 -> 0.10Hurt
Use LCRankNet for early termination instead of training every configuration to completionRandom NAS in the NASNet search space · QuickdrawTime (GPU hours): 1045 -> 28Improved
Add metadata from other datasets (ablation)LCRankNet · Learning-curve ranking across image datasetsSpearman rank correlationImproved
Add architecture description to the learning-curve representation (ablation)LCRankNet · Learning-curve rankingSpearman rank correlationImproved
Add learning-curve information to the architecture-only model (ablation)LCRankNet · Learning-curve rankingSpearman rank correlationImproved
Use architecture information alone rather than the full representation (ablation)LCRankNet · Short learning curvesSpearman rank correlationMixed
Use learning-curve information alone rather than the full representation (ablation)LCRankNet · Learning-curve rankingSpearman rank correlationMixed
Replace pairwise ranking loss with pointwise L2 loss (ablation)LCRankNet · Short learning curvesSpearman rank correlationImproved
Replace pairwise ranking loss with pointwise L2 loss (ablation)LCRankNet · Longer learning curvesSpearman rank correlationHurt

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