Pattern recognition in high-energy physics

Abstract

Pattern recognition is of crucial importance to many high-energy physics experiments during their analysis phase. This review gives a short introduction to those aspects of experiments that require the application of pattern recognition methods, which are discussed in detail and illustrated with examples from high-energy physics experiments of the last 15 years. At the end, a number of papers are recommended dealing with track and vertex fitting.

Development of a pattern recognition algorithm for particle identification on the SciCRT in the Sierra Negra Volcano Summit

“…Pattern recognition has a wide range of applications, including speech recognition, electrocardiogram analysis, scene analysis and of course many high-energy physics experiments. In general, the pattern recognition can be defined as [4]:…”

“…A possible solution is the tree-search algorithm [4], this uses templates with different resolution, ordered in hierarchy. In others words, we can first match a small set of patterns with coarse resolution and in the next step, a second set of templates with a better resolution is possible.…”

High resolution multiparticle tracking without preprocessing via elastic tracking

“…Recently we proposed a new method [l], called Elastic Tracking (ET), to extend tracking capabilities to much higher track densities than previously possible with conventional methods [2,3,4]. The track density, ptrack, defined here by the ratio of the average distance between measured points along a track to the average distance between tracks provides a convenient measure of the complexity of a given tracking task.…”

“…where the density induced by track a is expressed as (3) with q(ra(r)) being proportional to the variable ionization charge density along the track. The finite range distribution, u(x), depends on the characteristics of the detector.…”

A real-time tracker for hadronic collider experiments

“…In this paper we propose the use of highly parallel dedicated processors to efficiently execute two fast track finding algorithms [ l , 21. The algorithms are based on the idea of a large bank of pre-calculated hit patterns to be compared to the event [3].…”