Dec 15 – 16, 2020
Centre for Particle Theory, Durham
Europe/London timezone

Machine Learning Surrogates for Rapid Dark Matter Direct Detection Calculations

Dec 15, 2020, 2:00 PM
Teams (Centre for Particle Theory, Durham)


Centre for Particle Theory, Durham

IPPP & Dept. Mathematical Sciences Durham University South Road Durham DH1 3LE


Dorian Amaral (IPPP, Durham University)


Dark matter direct detection is one of the most popular ways of searching for WIMP dark matter. However, the calculations involved when evaluating the recoil spectra can be expensive, especially when exploring the multi-dimensional parameter space resulting from the EFT description of dark matter and the inclusion of astrophysical uncertainties. Therefore, it is instructive to seek ways of accelerating this calculation. We explore the performance of several possible surrogate models which can replace the exact calculation, with a focus on machine learning models. We find that relatively simple surrogate models can adequately approximate the complete calculation with a sizeable jump in performance.

Length of talk 15-25 minutes
Would you be interested in receiving feedback on your talk? Yes
Will you be pre-recording your talk? No
Are you happy for your talk to be recorded? Yes

Primary author

Dorian Amaral (IPPP, Durham University)

Presentation materials