Table of Contents
Fetching ...

Techniques for improved heavy particle searches with jet substructure

Stephen D. Ellis, Christopher K. Vermilion, Jonathan R. Walsh

Abstract

We present a generic method for improving the effectiveness of heavy particle searches in hadronic channels at the Large Hadron Collider. By selectively removing, or pruning, protojets from the substructure provided by a k_T-style jet algorithm, we improve the mass resolution for heavy decays and decrease the QCD background. We show that the protojets removed are typical of soft radiation and underlying event contributions, and atypical of accurately reconstructed heavy particles.

Techniques for improved heavy particle searches with jet substructure

Abstract

We present a generic method for improving the effectiveness of heavy particle searches in hadronic channels at the Large Hadron Collider. By selectively removing, or pruning, protojets from the substructure provided by a k_T-style jet algorithm, we improve the mass resolution for heavy decays and decrease the QCD background. We show that the protojets removed are typical of soft radiation and underlying event contributions, and atypical of accurately reconstructed heavy particles.

Paper Structure

This paper contains 4 equations, 3 figures.

Figures (3)

  • Figure 1: Pruned and unpruned top (a) and $W$ (b) mass window widths (in GeV/$\text{c}^2$) versus $p_{T}$ window center (in GeV/c) for both $\text{k}_{\text{T}}$ and CA algorithms.
  • Figure 2: $S$ vs. $p_{T}$ for the CA and $\text{k}_{\text{T}}$ algorithms, using $D = 1.0$. Statistical errors, due to limited QCD sample sizes after cuts, are shown.
  • Figure 3: $S_{D}$ vs. $p_{T}$ for the CA and $\text{k}_{\text{T}}$ algorithms. The line at $S_{D}$ = 1 separates the regions where a tuned $D$ helps (above the line) and does not (below). The lowest $p_{T}$ bin is not shown because the $D$ value does not change ($S_{D} = 1$). Statistical errors are shown.