AI Paper Atlas

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

Information-Theoretic Characterization of the Generalization Error for Iterative Semi-Supervised Learning

Haiyun He, Hanshu Yan, Vincent Y. F. Tan

JMLR 2022 · Label-efficient and robust learning › Semi-supervised learning · approach: Holistic semi-supervised pipelines · 4 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
add iterative pseudo-labelled unlabelled data when classes are easy to distinguishIterative self-training with pseudo-labelling · CIFAR-10 horse-ship / automobile-truckgen-error: converges to around 0.25 after 5 iterationsImproved
add iterative pseudo-labelling on 10-class digitsResNet-6 self-training · MNISTgen-error: converges to around 0.25 after 20 iterationsImproved
add pseudo-labelled data for hard-to-distinguish classesResNet-10 self-training · CIFAR-10 cat-doggen-errorHurt
add l2 regularization with weight decay 0.0005 (ablation)ResNet-10 self-training · CIFAR-10 cat-dogtest accuracy: test accuracy improved by 0.6%Improved
increase weight decay (ablation)ResNet-10 self-training · CIFAR-10 cat-dogaverage gen-error over last 10 iterationsImproved
use all 9,500 unlabelled images in every iteration instead of sampling 2500 (ablation)ResNet-10 self-training · CIFAR-10 horse-shipgen-errorHurt
use unlabelled data when class variance is small (sigma^2=0.6^2)bGMM iterative SSL · Synthetic bGMM d=2, n=10, m=1000gen-error: decreases significantly from t=0 to t=1 then converges to a non-zero constantImproved
use unlabelled data when class variance is large (sigma^2=3^2)bGMM iterative SSL · Synthetic bGMM d=2, n=10, m=1000gen-errorHurt
add l2 regularization lambda = 0, 0.1, 0.5, 2 (ablation)bGMM iterative SSL · Synthetic bGMM sigma=3, d=2gen-errorImproved
reuse labelled data in every iteration (ablation)bGMM iterative SSL · Synthetic bGMM sigma=0.6, n=10, m=1000gen-errorNo clear effect
reduce unlabelled data from m=1000 to m=100 (ablation)bGMM iterative SSL · Synthetic bGMM with labelled data reusegen-error: m=100 higher than m=1000Hurt
add pseudo-labelled data when labelled set is large (n above 30)bGMM iterative SSL · Synthetic bGMM m=1000, sigma=0.6, d=2gen-error: gen_0 smaller than gen_1 for n larger than 30Hurt

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