# Vector Boson Fusion versus Gluon Fusion

## TL;DR

This paper introduces a new 2-step boosted-decision-tree technique to better distinguish vector-boson fusion events from gluon fusion contamination, enhancing the study of electroweak symmetry breaking in Higgs processes.

## Contribution

It develops a novel machine learning approach incorporating jet-shape variables to improve separation of VBF from ggF in Higgs analyses.

## Key findings

- The method significantly reduces gluon fusion contamination in VBF samples.
- Jet-shape variables like girth improve discrimination power.
- Enhanced environment for probing electroweak symmetry breaking.

## Abstract

Vector-boson fusion (VBF) is a clean probe of the electroweak-symmetry breaking (EWSB), which inevitably suffers from some level of contamination due to the gluon fusion (ggF). In addition to the jet variables used in the current experimental analysis, we analyze a few more jet-shape variables defined by the girth and integrated jet-shape. Taking $H\to W W^* \to e \nu \mu \nu$ and $H \to \gamma\gamma$ as examples, we perform the analysis with a new technique of 2-step boosted-decision-tree method, which significantly reduces the contamination of the ggF in the VBF sample, thus, providing a clean environment in probing the EWSB sector.