Bayesian model selection for electromagnetic kaon production on the nucleon

TL;DR

This paper uses Bayesian analysis with nested sampling to identify the most probable Regge model variants for electromagnetic kaon production on nucleons, providing a statistically rigorous model selection approach.

Contribution

It introduces a Bayesian evidence-based method for selecting the best Regge model variants in kaon photoproduction, surpassing traditional chi^2 minimization.

Findings

Decisive Bayesian evidence for a specific K+ Lambda model variant.

Inconclusive evidence for K+ Sigma0 model variants.

Bayesian approach outperforms chi^2 minimization in model selection.

Abstract

We present the results of a Bayesian analysis of a Regge model to describe the background contribution for K+ Lambda and K+ Sigma0 photoproduction. The model is based on the exchange of K+(494) and K*+(892) trajectories in the t-channel. We utilise the Bayesian evidence Z to determine the best model variant for each channel. The Bayesian evidence integrals were calculated using the Nested Sampling algorithm. For different prior widths, we find decisive Bayesian evidence (\Delta ln Z ~ 24) for a K+ Lambda photoproduction Regge model with a positive vector coupling and a negative tensor coupling constant for the K*+(892) trajectory, and a rotating phase factor for both trajectories. Using the chi^2 minimisation method, one could not draw this conclusion from the same dataset. For the K+ Sigma0 photoproduction Regge model, on the other hand, the difference between the evidence integrals is insufficient to pinpoint one model variant.