Multiple testing for signal-agnostic searches for new physics with machine learning
2025

Using Machine Learning to Find New Physics

publication Evidence: moderate

Author Information

Author(s): Grosso Gaia, Letizia Marco

Primary Institution: NSF AI Institute for Artificial Intelligence and Fundamental Interactions, Cambridge, MA USA

Hypothesis

How can multiple testing strategies enhance signal-agnostic searches for new physics?

Conclusion

Combining different tests improves the detection of new physics signals and provides a more uniform response to various anomalies.

Supporting Evidence

  • Combining different tests can achieve performances comparable to the best available test.
  • The proposed methodology is valid beyond machine learning approaches.
  • Multiple testing strategies can enhance model-agnostic searches for new physics.

Takeaway

This study shows that using different tests together can help scientists find new things in physics better than using just one test.

Methodology

The study explores multiple testing strategies in machine learning for signal-agnostic searches, focusing on the New Physics Learning Machine.

Potential Biases

Model selection can introduce bias towards specific families of new physics signals.

Digital Object Identifier (DOI)

10.1140/epjc/s10052-024-13722-5

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication