Work

Online-compatible unsupervised nonresonant anomaly detection

Year: 2022

Type: article

Abstract: The authors of this paper employ two (or more) autoencoders to provide a complete strategy for unsupervised non-resonant anomaly detection. Both signal extraction and data-driven background estimation can be determined with decorrelated autoencoders. The method shows strong performance on test datasets and has the advantage of being online-compatible.

Source: Physical review. D/Physical review. D.

DOI: 10.1103/physrevd.105.055006

Authors: Vinicius Mikuni, Benjamin Nachman, David Shih

Institutions: Lawrence Berkeley National Laboratory, National Energy Research Scientific Computing Center, University of California, Berkeley, Rutgers, The State University of New Jersey

Language: English

Topic: Anomaly Detection Techniques and Applications

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

Funders: U.S. Department of Energy, Office of Science