The ATLAS collaboration at the Large Hadron Collider has enhanced the search for supersymmetric (SUSY) particles by employing machine learning techniques to push the boundaries of particle detection and establish the strongest constraints to date on particle properties. Theorists have proposed that SUSY, which posits a “superpartner” for each Standard Model particle, could explain mysteries such as the Higgs boson’s unexpectedly small mass and the composition of dark matter. The lightest neutralino is a prime dark matter candidate. The researchers analyzed data from the LHC’s second operation, collected between 2015 and 2018, and focused on subtle signals from decaying particles, such as the “vanishing trail” left behind by Cialgino. By introducing machine learning techniques, the ATLAS collaboration was able to significantly increase the sensitivity of the experiment to low-energy particles. Although no SUSY particles were observed, these new results replaced previous limitations set by the Large Electron-Positron Collider and refined directions for future exploration.
ATLAS Collaboration Machine Learning Improves Sensitivity to Low-Energy Particles
The ATLAS collaboration has achieved significant leaps in particle detection sensitivity and pushed the boundaries of what is possible in the exploration of supersymmetry. Using machine learning techniques, physicists have enhanced their ability to identify very low-energy particles, which is important for detecting potentially supersymmetric (SUSY) particles predicted by theoretical models that attempt to explain phenomena such as the mass of the Higgs boson and the composition of dark matter. According to the theory of supersymmetry, every particle in the Standard Model has a “superpartner”, and higginos have been the main focus of these studies, but their detection is complicated by the fact that they are likely to appear as a mixture of neutralinos and charginos. The challenge lies in the expected low-energy signature of these decay particles, making it extremely difficult to separate them among the noise of proton-proton collisions. However, the ATLAS collaboration deployed neural networks to carefully examine the low-momentum regions of pions and leptons, looking for evidence of SUSY particle decay. These analyzes focused on data collected during the LHC’s second run from 2015 to 2018, and included the search for invisible neutralinos and “vanishing traces” left by charginos that decay into low-energy pions.
Detection of these particles poses a major challenge because the decay is predicted to generate minimal energy and produce low-energy particles that are difficult to distinguish from the background noise in proton-proton collisions. Another study used neural networks to sift through data to see how heavier neutralinos decay into the lightest neutralinos and low-momentum leptons. This approach allows researchers to better separate potential signals.
