New approach to the Parton Distribution Functions: Self-Organizing Maps

TL;DR

This paper introduces a novel PDF fitting method using Self-Organizing Maps that allows for interactive control and potentially reduces systematic bias in the determination of Parton Distribution Functions.

Contribution

It presents a new neural network-based fitting technique employing SOMs, offering increased user interaction and bias control compared to traditional methods.

Findings

SOMs effectively cluster PDF samples during fitting.

The method enables user-guided iterative refinement.

Potential for improved systematic bias management.

Abstract

We propose a Parton Distribution Function (PDF) fitting technique which is based on an interactive neural network algorithm using Self-Organizing Maps (SOMs). SOMs are visualization algorithms based on competitive learning among spatially-ordered neurons. Our SOMs are trained with stochastically generated PDF samples. On every optimization iteration the PDFs are clustered on the SOM according to a user-defined feature and the most promising candidates are selected as a seed for the subsequent iteration. Our main goal is thus to provide a fitting procedure that, at variance with the global analyses and standard neural network approaches, allows for an increased control of the systematic bias by enabling user interaction in the various stages of the fitting process.