Multiple testing for signal-agnostic searches for new physics with machine learning
In this work, we address the question of how to enhance signal-agnostic searches by leveraging multiple testing strategies. Specifically, we consider hypothesis tests relying on machine learning, where model selection can introduce a bias towards specific families of new physics signals. Focusing on...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Springer Berlin Heidelberg
2025
|
| Online Access: | https://hdl.handle.net/1721.1/157946 |
| _version_ | 1824457969703583744 |
|---|---|
| author | Grosso, Gaia Letizia, Marco |
| author2 | Massachusetts Institute of Technology. Laboratory for Nuclear Science |
| author_facet | Massachusetts Institute of Technology. Laboratory for Nuclear Science Grosso, Gaia Letizia, Marco |
| author_sort | Grosso, Gaia |
| collection | MIT |
| description | In this work, we address the question of how to enhance signal-agnostic searches by leveraging multiple testing strategies. Specifically, we consider hypothesis tests relying on machine learning, where model selection can introduce a bias towards specific families of new physics signals. Focusing on the New Physics Learning Machine, a methodology to perform a signal-agnostic likelihood-ratio test, we explore a number of approaches to multiple testing, such as combining p-values and aggregating test statistics. Our findings show that it is beneficial to combine different tests, characterised by distinct choices of hyperparameters, and that performances comparable to the best available test are generally achieved, while also providing a more uniform response to various types of anomalies. This study proposes a methodology that is valid beyond machine learning approaches and could in principle be applied to a larger class model-agnostic analyses based on hypothesis testing. |
| first_indexed | 2025-02-19T04:18:27Z |
| format | Article |
| id | mit-1721.1/157946 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2025-02-19T04:18:27Z |
| publishDate | 2025 |
| publisher | Springer Berlin Heidelberg |
| record_format | dspace |
| spelling | mit-1721.1/1579462025-01-07T03:54:06Z Multiple testing for signal-agnostic searches for new physics with machine learning Grosso, Gaia Letizia, Marco Massachusetts Institute of Technology. Laboratory for Nuclear Science In this work, we address the question of how to enhance signal-agnostic searches by leveraging multiple testing strategies. Specifically, we consider hypothesis tests relying on machine learning, where model selection can introduce a bias towards specific families of new physics signals. Focusing on the New Physics Learning Machine, a methodology to perform a signal-agnostic likelihood-ratio test, we explore a number of approaches to multiple testing, such as combining p-values and aggregating test statistics. Our findings show that it is beneficial to combine different tests, characterised by distinct choices of hyperparameters, and that performances comparable to the best available test are generally achieved, while also providing a more uniform response to various types of anomalies. This study proposes a methodology that is valid beyond machine learning approaches and could in principle be applied to a larger class model-agnostic analyses based on hypothesis testing. 2025-01-06T18:37:58Z 2025-01-06T18:37:58Z 2025-01-04 2025-01-05T04:11:59Z Article http://purl.org/eprint/type/JournalArticle https://hdl.handle.net/1721.1/157946 Grosso, G., Letizia, M. Multiple testing for signal-agnostic searches for new physics with machine learning. Eur. Phys. J. C 85, 4 (2025). PUBLISHER_CC en https://doi.org/10.1140/epjc/s10052-024-13722-5 The European Physical Journal C Creative Commons Attribution https://creativecommons.org/licenses/by/4.0/ The Author(s) application/pdf Springer Berlin Heidelberg Springer Berlin Heidelberg |
| spellingShingle | Grosso, Gaia Letizia, Marco Multiple testing for signal-agnostic searches for new physics with machine learning |
| title | Multiple testing for signal-agnostic searches for new physics with machine learning |
| title_full | Multiple testing for signal-agnostic searches for new physics with machine learning |
| title_fullStr | Multiple testing for signal-agnostic searches for new physics with machine learning |
| title_full_unstemmed | Multiple testing for signal-agnostic searches for new physics with machine learning |
| title_short | Multiple testing for signal-agnostic searches for new physics with machine learning |
| title_sort | multiple testing for signal agnostic searches for new physics with machine learning |
| url | https://hdl.handle.net/1721.1/157946 |
| work_keys_str_mv | AT grossogaia multipletestingforsignalagnosticsearchesfornewphysicswithmachinelearning AT letiziamarco multipletestingforsignalagnosticsearchesfornewphysicswithmachinelearning |