SciPost Submission Page
Anomaly Awareness
by Charanjit K. Khosa, Veronica Sanz
This Submission thread is now published as
Submission summary
| Submission information |
| Preprint Link: |
https://arxiv.org/abs/2007.14462v4
(pdf)
|
| Date accepted: |
2023-06-01 |
| Date submitted: |
2023-03-14 17:22 |
| Submitted by: |
Khosa, Charanjit Kaur |
| Submitted to: |
SciPost Physics |
| Ontological classification |
| Academic field: |
Physics |
| Specialties: |
- High-Energy Physics - Phenomenology
|
| Approach: |
Phenomenological |
Abstract
We present a new algorithm for anomaly detection called Anomaly Awareness. The algorithm learns about normal events while being made aware of the anomalies through a modification of the cost function. We show how this method works in different Particle Physics situations and in standard Computer Vision tasks. For example, we apply the method to images from a Fat Jet topology generated by Standard Model Top and QCD events, and test it against an array of new physics scenarios, including Higgs production with EFT effects and resonances decaying into two, three or four subjets. We find that the algorithm is effective identifying anomalies not seen before, and becomes robust as we make it aware of a varied-enough set of anomalies.
Author comments upon resubmission
We would like to thank the referees for their final comments. We have addressed the two remaining issues: 1.) now the former .png figures have been substituted by higher quality .pdfs and 2.) we have compared the sensitivity to the EFT anomaly from the semi-supervised AA task with a supervised classifier trained on SM background vs EFT. We have obtained roughly a factor 2 of difference in sensitivity. We have added a paragraph on this comparison at the end of the section 'Anomaly Detection'. We hope that with these final modifications, the paper is ready for publication.
