Research map / Label-efficient and robust learning / Recognition and long-tail
Conditional Bernoulli Mixtures for Multi-label Classification
Cheng Li, Bingyu Wang, Virgil Pavlu, Javed Aslam
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use CBM mixture instead of independent binary relevance (LR) | logistic regression · MEDIAMILL | subset accuracy: 9.6 -> 13.5 | Improved |
| Use CBM instead of BinRel (LR) | logistic regression · SCENE | subset accuracy: 51.5 -> 69.7 | Improved |
| Use CBM instead of PowSet (LR) | logistic regression · RCV1 | subset accuracy: 50.2 -> 49.9 | Hurt |
| Use CBM instead of PowSet (LR) | logistic regression · MEDIAMILL | subset accuracy: 9.0 -> 13.5 | Improved |
| Swap LR base learner for gradient boosted trees in CBM | CBM · MEDIAMILL | subset accuracy: 13.5 -> 14.1 | Improved |
| Swap LR base learner for gradient boosted trees in CBM | CBM · SCENE | subset accuracy: 69.7 -> 70.5 | Improved |
| Swap LR base learner for GB in CBM | CBM · RCV1 | subset accuracy: 49.9 -> 43.0 | Hurt |
| Swap LR for GB in BinRel | BinRel · RCV1 | subset accuracy: 40.4 -> 30.1 | Hurt |
| Swap LR for GB in BinRel | BinRel · SCENE | subset accuracy: 51.5 -> 59.3 | Improved |
| Swap LR for GB in PowSet | PowSet · TMC2007 | subset accuracy: 28.2 -> 23.1 | Hurt |
| Increase number of mixture components K from 1 to 15 (ablation) | CBM · TMC | subset accuracy | Improved |
| Increase K beyond about 30 (ablation) | CBM · TMC | subset accuracy | No clear effect |
Closest papers by meaning
- LCBM: A Multi-View Probabilistic Model for Multi-Label Classification (TPAMI 2020 · similarity 0.89)
- Towards Reliable Multi-Label Classification via Conditional Dependency Modeling (ECCV 2026 · similarity 0.87)
- Multi-Label Ranking From Positive and Unlabeled Data (CVPR 2016 · similarity 0.87)
- A Gaussian Process-Bayesian Bernoulli Mixture Model for Multi-Label Active Learning (NeurIPS 2021 · similarity 0.87)
- Consistent algorithms for multi-label classification with macro-at-$k$ metrics (ICLR 2024 · similarity 0.86)
- Taming the Sigmoid Bottleneck: Provably Argmaxable Sparse Multi-Label Classification (AAAI 2024 · similarity 0.85)
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.
