AI Paper Atlas

Research map / Label-efficient and robust learning / OOD and anomaly detection

ADGym: Design Choices for Deep Anomaly Detection

Minqi Jiang, Chaochuan Hou, Ao Zheng, Songqiao Han, Hailiang Huang, Qingsong Wen, Xiyang Hu, Yue Zhao

NeurIPS 2023 · Label-efficient and robust learning › OOD and anomaly detection · approach: Deep one-class and tabular anomaly detection · 5 citations (OpenAlex)

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 changedWhereMeasuredEffect
Select pipelines with a meta-predictor instead of using the strongest fixed SOTA methodADGym · Real-world datasets, n_a=5AUCROC: 0.784 -> 0.821Improved
Select pipelines with a meta-predictor instead of using the strongest fixed SOTA methodADGym · Real-world datasets, n_a=10AUCROC: 0.816 -> 0.849Improved
Select pipelines with a meta-predictor instead of using the strongest fixed SOTA methodADGym · Real-world datasets, n_a=20AUCROC: 0.843 -> 0.869Improved
Use supervised validation selection rather than random pipeline selectionADGym · Real-world datasets, n_a=5AUCROC: 0.735 -> 0.613Hurt
Ensemble predicted pipelines rather than using a single DL meta-predictor (ablation)ADGym DL meta-predictor · Real-world datasets, n_a=5AUCROC: 0.800 -> 0.813Improved
Ensemble predicted pipelines rather than using a single DL meta-predictor (ablation)ADGym DL meta-predictor · Real-world datasets, n_a=10AUCROC: 0.836 -> 0.843Improved
Ensemble predicted pipelines rather than using a single DL meta-predictor (ablation)ADGym DL meta-predictor · Real-world datasets, n_a=20AUCROC: 0.859 -> 0.865Improved
Expand the candidate design space (ablation)ADGym meta-predictors · Real-world datasets, n_a=5AUCROC: 0.813 -> 0.829Improved
Expand the candidate design space (ablation)ADGym meta-predictors · Real-world datasets, n_a=10AUCROC: 0.843 -> 0.853Improved
Expand the candidate design space (ablation)ADGym meta-predictors · Real-world datasets, n_a=20AUCROC: 0.865 -> 0.876Improved
Refine the design space by removing choices below the training-set median (ablation)ADGym meta-predictors · Real-world datasetsAUCROCNo clear effect
Use more complicated predictor losses instead of MSE (ablation)ADGym meta-predictor · Real-world datasetsAUCROCNo clear effect

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy