New algorithm simplifies how physicists sort particle collision data
Researchers have developed a spectral clustering method that outperforms the industry-standard algorithm for reconstructing particle jets from high-energy collisions—without requiring manual tuning for different experiments. The advance could accelerate physics discoveries at major facilities like the Large Hadron Collider and reduce computational complexity in analyzing collision data.
Originaltitel: Spectral clustering for jet reconstruction
<p>We present a new approach to jet definition alternative to clustering methods, such as the anti-k<sub>T</sub> scheme, that exploit kinematic data directly. Instead the new method uses kinematic information to represent the particles in a multidimensional space, as in spectral clustering. After confirming its Infra-Red (IR) safety, we compare its performance in analysing gg→H<sub>125 GeV</sub>→H<sub>40 GeV</sub>H<sub>40 GeV</sub>→bb¯bb¯, gg→H<sub>500 GeV</sub>→H<sub>125 GeV</sub>H<sub>125 GeV</sub>→bb¯bb¯ and gg,qq¯→tt¯→bb¯W+W−→bb¯jjlνl events from Monte Carlo (MC) samples, specifically, in reconstructing the relevant final states, to that of the anti-k<sub>T</sub> algorithm. Finally, we show that the results for spectral clustering are obtained without any change in the parameter settings of the algorithm, unlike the anti-k<sub>T</sub> case, which requires the cone size to be adjusted to the physics process under study.</p>