AI Paper Atlas

Research map / Label-efficient and robust learning / Semi-supervised learning

Not All Unlabeled Data are Equal: Learning to Weight Data in Semi-supervised Learning

Zhongzheng Ren, Raymond Yeh, Alexander Schwing

NeurIPS 2020 · Label-efficient and robust learning › Semi-supervised learning · approach: Safe and realistic semi-supervised learning · 20 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
Learn per-example weights instead of using a shared UDA weightUDA · CIFAR-10 and SVHN labeled-data splitstest error rateImproved
Learn per-example weights instead of using a shared FixMatch weightFixMatch (CTA) · CIFAR-10 and SVHN reported splitstest error rateImproved
Learn per-example weights for UDAUDA with BERT · IMDb, 20 labeled examples, sequence length 128test error rate: 8.98$\pm$0.26 -> 8.51$\pm$0.14Improved
Use per-example weights rather than tune a single shared weight (ablation)Proposed method with UDA · CIFAR-10 labeled-data splitstest error rateImproved
Use masked Adam updates rather than vanilla Adam (ablation)Per-example weight optimizer · CIFAR-10, UDAvalidation error rateImproved
Use masked Adam updates rather than SGD (ablation)Per-example weight optimizer · CIFAR-10, UDAvalidation error rateImproved
Increase validation-set size (ablation)Proposed method with UDA · CIFAR-10, 250 and 4000 labeled examplestest error rateImproved
Set inner steps N to 30 rather than 100 (ablation)Proposed method with UDA · CIFAR-10, 4000 labeled examplesvalidation error rate: 3.42 -> 5.13Hurt
Set inner steps N to 300 rather than 100 (ablation)Proposed method with UDA · CIFAR-10, 4000 labeled examplesvalidation error rate: 3.42 -> 4.59Hurt
Set outer-loop step size eta to 0.1 rather than 0.01 (ablation)Proposed method with UDA · CIFAR-10, 4000 labeled examplesvalidation error rate: 3.42 -> 6.16Hurt
Set outer-loop step size eta to 0.001 rather than 0.01 (ablation)Proposed method with UDA · CIFAR-10, 4000 labeled examplesvalidation error rate: 3.42 -> 4.10Hurt
Use the exact inverse Hessian of only the last layer for influence estimation (ablation)Proposed method with UDA · CIFAR-10test error rate: 4.51% and 4.85% -> 4.43%Improved

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