K-matrix parametrized differential cross section datasets for deep neural network feature extraction

Abstract

Enhancements in the scattering amplitude are usually associated with resonances. Numerous resonances are detected in experiments, with wide, overlapping, and intricate structures, complicating their identification. To investigate the properties of these resonances, the invariant mass spectrum of the reactions is investigated. Most often in practice, the angular momentum state is of interest. This requires the extraction of different angular momentum contributions and the associated resonance parameters from the scattering amplitude of the process. In hadron spectroscopy, this can be accomplished with a method called partial wave analysis (PWA). In this paper, K-matrix formalism is applied to the parametrization of the scattering amplitude. To account for the angular dependence, the Blatt-Weisskopf barrier factors are incorporated into the formalism. Differential cross sections from the parametrization are generated in various parameter variation schemes. The overlapping features and mixed behaviors observed across the plots complicate visual analysis, making qualitative conclusions based on inspection unreliable. To address this complexity, the study proposed using a deep neural network (DNN) approach in the next phase of analysis.