Research map / Label-efficient and robust learning / Noisy labels
Learning with Marginalized Corrupted Features and Labels Together
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| add marginalized label denoising (autoencoder) regularization to marginalized corrupted features | MCF (Blankout) · Bibtex | F1: 0.3671 -> 0.4125 | Helped |
| add marginalized corrupted features (blankout) to FastTag label enrichment | FastTag · Bookmarks | F1: 0.2492 -> 0.2693 | Helped |
| add marginalized corrupted features (blankout) to FastTag label enrichment | FastTag · Enron | F1: 0.2342 -> 0.2524 | Helped |
| use blankout rather than Poisson corruption for features (ablation) | RMCV · Bibtex | F1: 0.3898 -> 0.4125 | Helped |
| add marginalized corrupted features (vs least squares) (ablation) | LeastSquares · Bibtex | F1: 0.3326 -> 0.3671 | Helped |
| RMCV vs MCF recall | MCF (Poisson) · Bibtex | recall: 0.5240 -> 0.4274 | Hurt |
| increase blankout corruption level q from low to high (~0.8) then beyond (ablation) | RMCV (Blankout) · Bibtex | F1 | Mixed |
| increase maximum tags per training document (ablation) | RMCV (Blankout) · Bibtex | F1 | Helped |
Closest papers by meaning
- Learning With Incomplete Labels (AAAI 2018 · similarity 0.88)
- Learning with Feature Network and Label Network Simultaneously (AAAI 2017 · similarity 0.86)
- Adaptive Integration of Partial Label Learning and Negative Learning for Enhanced Noisy Label Learning (AAAI 2024 · similarity 0.83)
- CGMatch: A Different Perspective of Semi-supervised Learning (CVPR 2025 · similarity 0.83)
- Partial Multi-View Multi-Label Classification via Semantic Invariance Learning and Prototype Modeling (ICML 2024 · similarity 0.82)
- Generalization Analysis for Multi-Label Learning (ICML 2024 · similarity 0.82)
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.