Research map / Label-efficient and robust learning / Recognition and long-tail
Automatic Unsupervised Ensemble Outlier Model Selection
Hong-Phuc Phan, Tuan-Anh Vu, Tung Kieu, Son Ha Xuan, Bin Yang, Christian S. Jensen
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 |
|---|---|---|---|
| add family-risk regularization (ablation) | MetaEns · 39 outlier detection datasets (ELECT benchmark) | Average Precision: 0.3995 -> 0.4308 | Improved |
| add diversity discount (ablation) | MetaEns · 39 outlier detection datasets (ELECT benchmark) | Average Precision: 0.4185 -> 0.4308 | Improved |
| use two-part classifier-regressor gain model instead of a single predictor (ablation) | MetaEns · 39 outlier detection datasets (ELECT benchmark) | Average Precision: 0.4133 -> 0.4308 | Improved |
| add stage-dependent acceptance gates (ablation) | MetaEns · 39 outlier detection datasets (ELECT benchmark) | Average Precision: 0.4257 -> 0.4308 | Improved |
| replace ELECT Top-10 aggregation with sequential MetaEns selection | ELECT primary detector · 39 outlier detection datasets (ELECT benchmark) | Average Precision: 0.4117 -> 0.4308 | Improved |
| add diversity discount (ablation) | MetaEns · 39 outlier detection datasets (ELECT benchmark) | AP rank: 77.0 -> 59.3 | Improved |
| add family-risk regularization (ablation) | MetaEns · 39 outlier detection datasets (ELECT benchmark) | AP rank: 72.0 -> 59.3 | Improved |
| use ExtraTrees as meta-model (ablation) | MetaEns · 39 outlier detection datasets (ELECT benchmark) | Average Precision | Improved |
| use oracle greedy rollouts instead of epsilon-greedy exploration for meta-training (ablation) | MetaEns · 39 outlier detection datasets (ELECT benchmark) | Average Precision | Improved |
| average all 297 detectors instead of selecting | Mega Ensemble · 39 outlier detection datasets (ELECT benchmark) | Average Precision: 0.3970 -> 0.4308 | Hurt |
| replace deep detector LUNAR with MetaEns on image/text transfer | ADBench image/text (20 datasets) · ADBench image/text (20 datasets) | Average Precision: +0.0197 AP | Improved |
Closest papers by meaning
- ALTBI: Constructing Improved Outlier Detection Models via Optimization of Inlier-Memorization Effect (AAAI 2025 · similarity 0.86)
- DOS: Diverse Outlier Sampling for Out-of-Distribution Detection (ICLR 2024 · similarity 0.85)
- ODIM: Outlier Detection via Likelihood of Under-Fitted Generative Models (ICML 2024 · similarity 0.85)
- SEAD: Unsupervised Ensemble of Streaming Anomaly Detectors (ICML 2025 · similarity 0.85)
- Out-Of-Distribution Detection with Diversification (Provably) (NeurIPS 2024 · similarity 0.85)
- Quantile-Based Maximum Likelihood Training for Outlier Detection (AAAI 2024 · similarity 0.84)
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.
