Research map / Learning theory and algorithms / Prediction and decision losses
PAC Guarantees and Effective Algorithms for Detecting Novel Categories
Si Liu, Risheek Garrepalli, Dan Hendrycks, Alan Fern, Debashis Mondal, Thomas G. Dietterich
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 |
|---|---|---|---|
| add isotonization and clipping of the estimated alien CDF (ablation) | Algorithm 1 · Synthetic 9-D normal, alpha=0.01, n=100 | recall: 0.710±0.033 -> 0.929±0.029 | Improved |
| add isotonization and clipping of the estimated alien CDF (ablation) | Algorithm 1 · Synthetic 9-D normal, alpha=0.01, n=100 | false positive rate: 0.033±0.027 -> 0.512±0.080 | Hurt |
| add isotonization and clipping of the estimated alien CDF (ablation) | Algorithm 1 · Synthetic 9-D normal, alpha=0.5, n=10000 | false positive rate: 0.036±0.001 -> 0.047±0.016 | Hurt |
| add isotonization and clipping of the estimated alien CDF (ablation) | Algorithm 1 · Synthetic and benchmark data | overall performance | No clear effect |
| increase sample size n from 100 to 10000 (ablation) | Algorithm 1 basic CDF · Synthetic 9-D normal, alpha=0.01 | recall: 0.710±0.033 -> 0.932±0.009 | Improved |
| increase alien proportion alpha from 0.01 to 0.5 (ablation) | Algorithm 1 basic CDF · Synthetic 9-D normal, n=100 | recall: 0.710±0.033 -> 0.970±0.005 | Improved |
| increase alien proportion alpha from 0.01 to 0.5 (ablation) | Algorithm 1 basic CDF · Synthetic 9-D normal, n=100 | false positive rate: 0.033±0.027 -> 0.156±0.036 | Hurt |
| overestimate alpha with alpha' = alpha + xi (ablation) | Algorithm 1 · UCI datasets | false positive rate: 0 to 0.246 | Hurt |
| overestimate alpha with alpha' = alpha + xi (ablation) | Algorithm 1 · UCI datasets | recall: 0.001 to 0.047 | Improved |
| estimate alpha from anomaly scores before RKHS embedding instead of raw data (ablation) | KM1/KM2 alpha estimators · Synthetic 9-D normal | alpha estimate bias | Improved |
| use SPY estimator for alpha instead of bt-patrasen, C-PS or C-ROC (ablation) | Algorithm 1 with alpha estimation · UCI datasets | false positive rate | Hurt |
| use iForest scores on covertype for bt-patrasen alpha estimate (ablation) | bt-patrasen · covertype | alpha estimate | Hurt |
Closest papers by meaning
- Open Category Detection with PAC Guarantees (ICML 2018 · similarity 0.98)
- Anomaly Heterogeneity Learning for Open-set Supervised Anomaly Detection (CVPR 2024 · similarity 0.83)
- Beyond Normal References: Discriminative Few-Shot Anomaly Detection (NeurIPS 2026 · similarity 0.82)
- OV-DQUO: Open-Vocabulary DETR with Denoising Text Query Training and Open-World Unknown Objects Supervision (AAAI 2025 · similarity 0.82)
- SetAD: Semi-Supervised Anomaly Learning in Contextual Sets (NeurIPS 2026 · similarity 0.81)
- VL-SAM-V2: Open-World Object Detection with General and Specific Query Fusion (NeurIPS 2025 · similarity 0.81)
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.
