Improving HPGe Detector Time Resolution Using Machine Learning

Hagen, Martin (2024) Improving HPGe Detector Time Resolution Using Machine Learning. Bachelor's Thesis, Physics.

Preview	Text BachelorMLThesis-final.pdf Download (3MB) | Preview
	Text toestemming_ Martin Hagen _ degree programme_ Physics.pdf Restricted to Registered users only Download (129kB)

Abstract

In this thesis, machine learning is applied to HPGe detector waveforms to cluster them with the goal of determining correlations that serve to improve detector time resolution. HPGe detectors are used in neutron cross section determination experiments; improvement of the moderate time resolution HPGe detectors possess is salient due to the neutron time-of-flight being closely linked to its energy, and in turn the reaction cross section. Close determination of neutron cross sections is important, as they play roles in nuclear reactor physics and design, astrophysics, radiation shielding and more. To investigate these correlations, the clustering algorithms KMeans and Hierarchical Clustering Algorithm (HCA) were applied to a version of the dataset reduced through principal component analysis. It was found that there was a strong correlation between the cluster index, corresponding to mean value of the first principal component P C1, and the time difference compared to a reference lanthanum bromide detector dT. Methods to apply corrections using clustering, as well as principal components analysis directly, are suggested and discussed.

Item Type:	Thesis (Bachelor's Thesis)
Supervisor name:	Kavatsyuk, M. and Kalantar-Nayestanaki, N.
Degree programme:	Physics
Thesis type:	Bachelor's Thesis
Language:	English
Date Deposited:	16 Jul 2024 11:52
Last Modified:	16 Jul 2024 11:52
URI:	https://fse.studenttheses.ub.rug.nl/id/eprint/33420

Actions (login required)

View Item