EE Seminar: Unsupervised Tabular Anomaly Detection with Diffusion Models (TabADM)
https://thetaray.zoom.us/j/3405403286?omn=89588930409&jst=2
ID: 3405403286
Electrical Engineering Systems ZOOM Seminar
Speaker: Guy Zamberg
M.Sc. student under the supervision of Prof. Amir Averbuch
Sunday, 27th July 2025, at 15:00
Unsupervised Tabular Anomaly Detection with Diffusion Models (TabADM)
Abstract
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 down-stream analysis.
In this work, we only assume access to contaminated data and present a diffusion-based probabilistic model effective for unsupervised anomaly detection.
Our model is trained to learn the density of normal samples by utilizing a unique rejection scheme to attenuate the influence of anomalies on the density estimation. At inference, we identify anomalies as samples in low-density regions.
We use real data to demonstrate that our method improves detection capabilities over baselines. Furthermore, our method is relatively stable to the dimension of the data and does not require extensive hyperparameter tuning.
השתתפות בסמינר תיתן קרדיט שמיעה = עפ"י רישום בצ'ט של שם מלא + מספר ת.ז.
