TabADM: Unsupervised Tabular Anomaly Detection with Diffusion Models

Guy Zamberg, Moshe Salhov, Ofir Lindenbaum +1

Tables are an abundant form of data with use cases across all scientific fields. Real-world datasets often contain anomalous samples that can negatively affect downstream analysis.…