Research map / Label-efficient and robust learning / Recognition and long-tail
Automatic Unsupervised Outlier Model Selection
Yue Zhao, Ryan Rossi, Leman Akoglu
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 |
|---|---|---|---|
| meta-learned model selection (MetaOD) vs always using iForest | outlier detection · POC | MAP: 0.1536 -> 0.2035 | Improved |
| MetaOD vs always using iForest | outlier detection · ST | MAP: 0.3197 -> 0.3382 | Improved |
| MetaOD selection vs ensemble of all models (ME) | outlier detection · POC | MAP: 0.1382 -> 0.2035 | Improved |
| fix U at meta-feature embedding (MetaOD_F) instead of learning U (ablation) | MetaOD · ST | MAP: 0.3382 -> 0.2594 | Hurt |
| concatenate P and meta-features (MetaOD_C) instead of regression approach (ablation) | MetaOD · ST | MAP: 0.3382 -> 0.1946 | Hurt |
| MetaOD vs MetaOD_C (ablation) | MetaOD · POC | MAP: 0.1980 -> 0.2035 | No clear effect |
| meta-feature initialization and cyclical learning rate for training stability | MetaOD · ST | MAP: 0.2981 -> 0.3382 | Improved |
| learned meta-learner (MetaOD) vs global best (GB) | model selection · POC | MAP: 0.1524 -> 0.2035 | Improved |
| local similarity meta-learners vs global best when task similarity is low | AS, ISAC vs GB · ST | MAP: 0.3137 vs 0.2704 | Mixed |
Closest papers by meaning
- ODIM: Outlier Detection via Likelihood of Under-Fitted Generative Models (ICML 2024 · similarity 0.88)
- ALTBI: Constructing Improved Outlier Detection Models via Optimization of Inlier-Memorization Effect (AAAI 2025 · similarity 0.87)
- How Does Unlabeled Data Provably Help Out-of-Distribution Detection? (ICLR 2024 · similarity 0.87)
- MetaOOD: Automatic Selection of OOD Detection Models (ICLR 2025 · similarity 0.87)
- DOS: Diverse Outlier Sampling for Out-of-Distribution Detection (ICLR 2024 · similarity 0.86)
- FlexUOD: The Answer to Real-world Unsupervised Image Outlier Detection (CVPR 2025 · similarity 0.86)
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.
