Research map / Optimization / Optimizers
Multivariate Triangular Quantile Maps for Novelty Detection
Jingjing Wang, Sun Sun, Yaoliang Yu
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 |
|---|---|---|---|
| train autoencoder and flow jointly with MGDA instead of freezing autoencoder after pretraining (Fix-) (ablation) | Ours-NLL · MNIST | AUROC: 0.9699 -> 0.9728 | Improved |
| joint MGDA training instead of pretrain then joint with fixed lambda 0.5 (ablation) | Ours-NLL · MNIST | AUROC: 0.9703 -> 0.9728 | Improved |
| joint MGDA training instead of fixed two-stage (Fix-) (ablation) | Ours-TQM2 · MNIST | AUROC: 0.9634 -> 0.9699 | Improved |
| MGDA-optimized lambda instead of fixed lambda 0.5 (ablation) | Ours-NLL · MNIST | AUROC: 0.9537 -> 0.9728 | Improved |
| MGDA-optimized lambda instead of best fixed lambda 0.99 (ablation) | Ours-NLL · MNIST | AUROC: 0.9729 -> 0.9728 | No clear effect |
| use density (NLL) scoring instead of 1-norm quantile (TQM1) scoring (ablation) | Ours · MNIST | AUROC: 0.9666 -> 0.9728 | Improved |
| use quantile infinity-norm scoring instead of density scoring (ablation) | Ours · MNIST | AUROC: 0.9728 -> 0.9531 | Hurt |
| add flow-based density/quantile scoring on latent instead of reconstruction error only (ablation) | Ours · KDDCUP | F1: 0.6296 -> 0.9622 | Improved |
| use TQM-infinity scoring instead of NLL scoring (ablation) | Ours · Thyroid | F1: 0.7312 -> 0.7527 | Improved |
| Ours-NLL vs GT, GPND on MNIST | Ours · MNIST | AUROC | Mixed |
Closest papers by meaning
- Latent Space Autoregression for Novelty Detection (CVPR 2019 · similarity 0.86)
- Gradient-Based Novelty Detection Boosted by Self-Supervised Binary Classification (AAAI 2022 · similarity 0.86)
- Projection Regret: Reducing Background Bias for Novelty Detection via Diffusion Models (NeurIPS 2023 · similarity 0.85)
- An Interpretable Evaluation of Entropy-based Novelty of Generative Models (ICML 2024 · similarity 0.84)
- Universal Novelty Detection Through Adaptive Contrastive Learning (CVPR 2024 · similarity 0.84)
- Hierarchical Novelty Detection via Fine-Grained Evidence Allocation (ICML 2024 · similarity 0.82)
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.
