Ridge regression for minimizing the couplings of hyperon resonances in the $K^+ \Lambda$ photoproduction

TL;DR

This paper applies Ridge regression to the isobar model of $K^+ \\Lambda$ photoproduction, reducing hyperon resonance couplings to more physical values and improving model robustness against polarization data.

Contribution

It introduces Ridge regression into hyperon resonance coupling estimation, leading to more realistic coupling values without altering the resonance set.

Findings

Hyperon couplings are significantly reduced to physical values.

The model's robustness is improved with the Ridge regression technique.

Insights into the role of hyperon resonances in photoproduction are gained.

Abstract

We employed the isobar model for investigating the $K^+\Lambda$ photoproduction process. We paid special attention to the recent CLAS polarization data and enhanced the $\chi^2$ minimization by adding a penalty term. Without changing the set of included resonances used by the model, this technique known as Ridge regression leads to reduced couplings that in previous studies acquired unreasonably large values. As a result, we have arrived at a much more robust model with hyperon couplings which are reduced to more physical values. This model serves us to extract valuable information on the background to the $K^+\Lambda$ photoproduction and particularly on the role of various hyperon resonances. The set of the nucleon resonances is the same with respect to previous fits but their role may have changed due to different couplings which they acquire in the present fit.