Learning new physics efficiently with nonparametric methods
Abstract We present a machine learning approach for model-independent new physics searches. The corresponding algorithm is powered by recent large-scale implementations of kernel methods, nonparametric learning algorithms that can approximate any continuous function given enough data....
| Main Authors: | , , , , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Springer Berlin Heidelberg
2022
|
| Online Access: | https://hdl.handle.net/1721.1/145780 |
| _version_ | 1811097873876516864 |
|---|---|
| author | Letizia, Marco Losapio, Gianvito Rando, Marco Grosso, Gaia Wulzer, Andrea Pierini, Maurizio Zanetti, Marco Rosasco, Lorenzo |
| author2 | Center for Brains, Minds, and Machines |
| author_facet | Center for Brains, Minds, and Machines Letizia, Marco Losapio, Gianvito Rando, Marco Grosso, Gaia Wulzer, Andrea Pierini, Maurizio Zanetti, Marco Rosasco, Lorenzo |
| author_sort | Letizia, Marco |
| collection | MIT |
| description | Abstract
We present a machine learning approach for model-independent new physics searches. The corresponding algorithm is powered by recent large-scale implementations of kernel methods, nonparametric learning algorithms that can approximate any continuous function given enough data. Based on the original proposal by D’Agnolo and Wulzer (Phys Rev D 99(1):015014, 2019,
arXiv:1806.02350
[hep-ph]), the model evaluates the compatibility between experimental data and a reference model, by implementing a hypothesis testing procedure based on the likelihood ratio. Model-independence is enforced by avoiding any prior assumption about the presence or shape of new physics components in the measurements. We show that our approach has dramatic advantages compared to neural network implementations in terms of training times and computational resources, while maintaining comparable performances. In particular, we conduct our tests on higher dimensional datasets, a step forward with respect to previous studies. |
| first_indexed | 2024-09-23T17:06:19Z |
| format | Article |
| id | mit-1721.1/145780 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T17:06:19Z |
| publishDate | 2022 |
| publisher | Springer Berlin Heidelberg |
| record_format | dspace |
| spelling | mit-1721.1/1457802023-06-28T20:16:45Z Learning new physics efficiently with nonparametric methods Letizia, Marco Losapio, Gianvito Rando, Marco Grosso, Gaia Wulzer, Andrea Pierini, Maurizio Zanetti, Marco Rosasco, Lorenzo Center for Brains, Minds, and Machines Abstract We present a machine learning approach for model-independent new physics searches. The corresponding algorithm is powered by recent large-scale implementations of kernel methods, nonparametric learning algorithms that can approximate any continuous function given enough data. Based on the original proposal by D’Agnolo and Wulzer (Phys Rev D 99(1):015014, 2019, arXiv:1806.02350 [hep-ph]), the model evaluates the compatibility between experimental data and a reference model, by implementing a hypothesis testing procedure based on the likelihood ratio. Model-independence is enforced by avoiding any prior assumption about the presence or shape of new physics components in the measurements. We show that our approach has dramatic advantages compared to neural network implementations in terms of training times and computational resources, while maintaining comparable performances. In particular, we conduct our tests on higher dimensional datasets, a step forward with respect to previous studies. 2022-10-11T17:59:11Z 2022-10-11T17:59:11Z 2022-10-05 2022-10-09T03:11:48Z Article http://purl.org/eprint/type/JournalArticle https://hdl.handle.net/1721.1/145780 The European Physical Journal C. 2022 Oct 05;82(10):879 PUBLISHER_CC en https://doi.org/10.1140/epjc/s10052-022-10830-y Creative Commons Attribution https://creativecommons.org/licenses/by/4.0/ The Author(s) application/pdf Springer Berlin Heidelberg Springer Berlin Heidelberg |
| spellingShingle | Letizia, Marco Losapio, Gianvito Rando, Marco Grosso, Gaia Wulzer, Andrea Pierini, Maurizio Zanetti, Marco Rosasco, Lorenzo Learning new physics efficiently with nonparametric methods |
| title | Learning new physics efficiently with nonparametric methods |
| title_full | Learning new physics efficiently with nonparametric methods |
| title_fullStr | Learning new physics efficiently with nonparametric methods |
| title_full_unstemmed | Learning new physics efficiently with nonparametric methods |
| title_short | Learning new physics efficiently with nonparametric methods |
| title_sort | learning new physics efficiently with nonparametric methods |
| url | https://hdl.handle.net/1721.1/145780 |
| work_keys_str_mv | AT letiziamarco learningnewphysicsefficientlywithnonparametricmethods AT losapiogianvito learningnewphysicsefficientlywithnonparametricmethods AT randomarco learningnewphysicsefficientlywithnonparametricmethods AT grossogaia learningnewphysicsefficientlywithnonparametricmethods AT wulzerandrea learningnewphysicsefficientlywithnonparametricmethods AT pierinimaurizio learningnewphysicsefficientlywithnonparametricmethods AT zanettimarco learningnewphysicsefficientlywithnonparametricmethods AT rosascolorenzo learningnewphysicsefficientlywithnonparametricmethods |