AI Paper Atlas

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

Exploiting Synthetically Generated Data with Semi-Supervised Learning for Small and Imbalanced Datasets

M. Peréz-Ortiz, P. Tiňo, R. Mantiuk, C. Hervás-Martínez

AAAI 2019 · Label-efficient and robust learning › Semi-supervised learning · approach: Confidence-based pseudo-label selection · 16 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 synthetic convex-combination patterns as unlabelled data in S³VMS³VM classification · Small high-dimensional datasetsaccuracy: 53.00 -> 66.00Improved
Use class-dependent synthetic oversampling in S³VM instead of class-independent oversampling (ablation)S³VM · Imbalanced UCI benchmark datasets, MCAR 0.8GM mean ranking: 3.65 -> 2.38Improved
Use an ensemble of S³VM models with different synthetic resampling distributions (ablation)S³VM · Small-sample UCI benchmark datasets, MCAR 0.5MAcc mean ranking: 3.54 -> 2.69Improved
Vary the number of generated synthetic patterns (ablation)S³VM-OvS · Synthetic data, d=100, N=50, V=0.5accuracyNo clear effect
Use real unlabelled data rather than synthetically generated unlabelled data (ablation)S³VM · Synthetic datasets, especially high-dimensional settingsaccuracyMixed
Apply class-independent instead of class-dependent synthetic oversampling (ablation)S³VM · Imbalanced benchmark datasetsMAcc mean ranking: 2.69 -> 3.58; 2.67 -> 3.67Hurt
Use transductive graph-based methods to label synthetic dataSynthetic-data semi-supervised classification · Small-sample benchmark datasetsmean ranking: 8.67 -> 3.21Hurt

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