Experimental Particle Physics

Dag Gillberg

Project 1: Jet calibration using the 'forward folding' technique. The most common fundamental particles produced at the Large Hadron Colliders are quarks and gluons. However, as a high energetic quark or gluon emerges from the collision point, it will undergo so-called hadronization and produce a collimated spray of particles known as a jet. Jets are hence extremely common at the Large Hadron Collider and very important for most of the physics analyses. This project is about investigating the potential of using a new technique called ‘forward folding’ to measure the ATLAS detector's response to jets. The two key quantities to be extracted are the mean jet detector response and the associated resolution. The forward folding technique is based on convolution. The aim of the project would be to investigate the use of this technique with ALTAS Z+jets data and simulation, which is something that has not been done before.

Project 2: Visualization and statistical compatibility tests of unbinned data. Most analyses at ATLAS that measure distributions provide and analyse these as histograms. However, new machine learning-based techniques have started to produce unbinned measurements, provided as event-datasets. Presenting such measurements as histograms does not do them justice: it would be better to construct a smooth function based on the unbinned data, which is possible using Gaussian Kernels, and perhaps other similar techniques. This project would be to investigate easy to visualize unbinned measurements, and quantify the uncertainties. In part it would include a literature review - looking if this problem already has been encountered in other fields. The goal would be to find a good way to visualize unbinned data for particle physics applications, and if times permit, investigate how to compare two different spectra to see if they are compatible or not within uncertainty.