Research map / Learning theory and algorithms / Prediction and decision losses
AutoMS: Automatic Model Selection for Novelty Detection with Error Rate Control
Yifan Zhang, Haiyan Jiang, Haojie Ren, Changliang Zou, Dejing Dou
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 AutoMS-JK model selection instead of METAOD | AutoMS-JK · Shuttle, target FDR 0.1 | empirical FDR: 0.573 -> 0.100 | Improved |
| use AutoMS-JK model selection instead of METAOD | AutoMS-JK · Shuttle, target FDR 0.1 | empirical TDR: 0.689 -> 0.942 | Improved |
| use AutoMS-JK instead of METAOD | AutoMS-JK · Covertype, target FDR 0.1 | empirical TDR: 0.960 -> 0.933 | Hurt |
| use AutoMS-JK instead of METAOD | AutoMS-JK · Credit Card, target FDR 0.1 | empirical TDR: 0.856 -> 0.758 | Hurt |
| select detector with AutoMS-JK instead of best single SRS detector | AutoMS-JK · Credit Card, target FDR 0.1 | empirical TDR: 0.674 -> 0.758 | Improved |
| select detector with AutoMS-JK instead of best single SRS detector | AutoMS-JK · Covertype, target FDR 0.1 | empirical TDR: 0.890 -> 0.933 | Improved |
| select detector with AutoMS-JK instead of best single SRS detector | AutoMS-JK · Satellite, target FDR 0.1 | empirical TDR: 0.963 -> 0.968 | No clear effect |
| use Jackknife prediction instead of split random sampling (SRS) inside AutoMS (ablation) | AutoMS · synthetic data, target FDR 0.1 | TDR | Improved |
| use Jackknife prediction instead of SRS inside AutoMS (ablation) | AutoMS · Shuttle, target FDR 0.05 | empirical TDR: 0.659 -> 0.883 | Improved |
| use Jackknife prediction instead of SRS inside AutoMS (ablation) | AutoMS · Satellite, target FDR 0.1 | empirical FDR: 0.134 -> 0.114 | Improved |
| raise target FDR level alpha from 0.05 to 0.20 | AutoMS-JK · Shuttle | empirical TDR: 0.883 -> 0.953 | Improved |
| choose among detectors and hyperparameters (model selection) rather than fixing one SRS detector | AutoMS · synthetic data | TDR | Improved |
Closest papers by meaning
- AutoSciDACT: Automated Scientific Discovery through Contrastive Embedding and Hypothesis Testing (NeurIPS 2025 · similarity 0.88)
- Latent Space Autoregression for Novelty Detection (CVPR 2019 · similarity 0.88)
- Universal Novelty Detection Through Adaptive Contrastive Learning (CVPR 2024 · similarity 0.86)
- Fake It Until You Make It : Towards Accurate Near-Distribution Novelty Detection (ICLR 2023 · similarity 0.85)
- Gradient-Based Novelty Detection Boosted by Self-Supervised Binary Classification (AAAI 2022 · similarity 0.85)
- ODIM: Outlier Detection via Likelihood of Under-Fitted Generative Models (ICML 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.
