AI Paper Atlas

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

A Neural Pre-Conditioning Active Learning Algorithm to Reduce Label Complexity

Seo Taek Kong, Soomin Jeon, Dongbin Na, Jaewon Lee, Hong-Seok Lee, Kyu-Hwan Jung

NeurIPS 2022 · Label-efficient and robust learning › Semi-supervised learning · approach: Safe and realistic semi-supervised learning · 0 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
replace BADGE with NPCactive learning + FixMatch · CIFAR-10, 50 labelsaccuracy (%): 86.43 ± 1.22 -> 94.63 ± 0.07Improved
replace BADGE with NPCactive learning + FixMatch · CIFAR-100, 300 labelsaccuracy (%): 24.55 ± 1.13 -> 40.92 ± 1.65Improved
replace random (passive) sampling with NPCactive learning + FixMatch · CIFAR-100, 700 labelsaccuracy (%): 52.69 ± 1.16 -> 55.76 ± 1.45Improved
replace ALBL with NPCactive learning + FixMatch · CIFAR-100, 500 labelsaccuracy (%): 49.88 ± 0.92 -> 48.16 ± 0.77Hurt
replace passive sampling with NPCactive learning + FixMatch · CIFAR-10, 30 labelsaccuracy (%): 78.08 ± 5.48 -> 85.09 ± 9.61, but Margin 87.36Mixed
use single-shot query model instead of zero-shot (random init) for BADGEBADGE · CIFAR-10, 60 labelsaccuracy (%): 94.42 ± 0.41 -> 70.28 ± 13.22Hurt
use single-shot query model instead of zero-shot for NPCNPC · CIFAR-10, 60 labelsaccuracy (%): 93.16 ± 2.34 -> 93.76 ± 0.58No clear effect
use NPC zero-shot (random init network) vs FixMatch on balanced data without DARPNPC · CIFAR-10accuracy (%): 86.19 -> 93.14 ± 1.42Improved
replace BADGE with NPC at zero-shot 40 labelsactive learning · CIFAR-10accuracy (%): 92.13 ± 3.24 -> 93.14 ± 1.42No clear effect
replace BADGE with NPC at single-shot 40 labelsactive learning · CIFAR-10accuracy (%): 86.70 ± 5.42 -> 92.26 ± 3.26Improved

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