AI Paper Atlas

Research map / Probabilistic and scientific modeling / Causal discovery

Causal-aware Anomaly Detection for Tabular Data

Dang Nguyen, Tu Anh Hoang Nguyen, Thuc Duy Le, Svetha Venkatesh, Trung Le, Sunil Gupta

ICML 2026 · Probabilistic and scientific modeling › Causal discovery · approach: Latent-variable causal estimation

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
replace the causal GAN with CTGAN in the CausalAno pipeline (ablation)CausalAno · All 28 datasets (average)AUC-ROC: 0.87 -> 0.65Hurt
replace the causal GAN with a standard GAN (ablation)CausalAno · All 28 datasets (average)AUC-ROC: 0.87 -> 0.61Hurt
use causal-aware discriminator features with Mahalanobis scoring versus best baselineCausalAno · 10 mixed-type datasets (average)AUC-ROC (%): 73.06 -> 76.76Improved
use CausalAno versus KNN and DRL on numeric-only dataCausalAno · 18 continuous-only datasets (average)AUC-ROC: 0.91 -> 0.93Improved
use CausalAno on causal-mechanism-violating synthetic anomaliesCausalAno · Synthetic Age-Edu-Salaryanomalies detected out of 1000: 710 -> 813 (vs DRL)Improved
contaminate training data with 0 to 5 percent anomalies (ablation)CausalAno · Annthyroid, OS, DAMRE, SMDAUC-ROCNo clear effect
vary causal coefficient lambda within 0.01 to 0.1 (ablation)CausalAno · sensitivity analysis (Appendix D)AUC-ROCNo clear effect
CausalAno versus LLM-based detectors on datasets with semantic columns (Lymp, SMD)CausalAno · Lymp and SMDAUC-ROC (%): Lymp 98.31 vs AnoLLM 99.63; SMD 93.69 vs LLM-DAS 95.74Mixed

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