The University of Southampton
University of Southampton Institutional Repository

Artificial Neural Network Processing of Double Electron-Electron Resonance Data

Artificial Neural Network Processing of Double Electron-Electron Resonance Data
Artificial Neural Network Processing of Double Electron-Electron Resonance Data
We show that artificial neural networks offer a powerful alternative to the established procedures for analysis of Double Electron-Electron Resonance (DEER) data. Simple five layer feed-forward neural networks were trained on a randomly generated synthetic data set. When the output of multiple such networks is combined to make an ensemble guess they are able to extract distance information with accuracy comparable to state-of-the-art methods. The variance in the ensemble output also provides a good estimation of the error in the result.

We then show that neural networks can be trained to predict both the exchange coupling interaction and the distance distribution in parallel. They were shown to provide a confident estimate at the magnitude of the exchange interaction while offering varying success when estimating the distance distribution.
Worswick, Steven Graham
e3f70075-e4f7-4b56-a20a-dbcfcecc90e9
Worswick, Steven Graham
e3f70075-e4f7-4b56-a20a-dbcfcecc90e9
Kuprov, Ilya
bb07f28a-5038-4524-8146-e3fc8344c065

Worswick, Steven Graham (2020) Artificial Neural Network Processing of Double Electron-Electron Resonance Data. Masters Thesis, 110pp.

Record type: Thesis (Masters)

Abstract

We show that artificial neural networks offer a powerful alternative to the established procedures for analysis of Double Electron-Electron Resonance (DEER) data. Simple five layer feed-forward neural networks were trained on a randomly generated synthetic data set. When the output of multiple such networks is combined to make an ensemble guess they are able to extract distance information with accuracy comparable to state-of-the-art methods. The variance in the ensemble output also provides a good estimation of the error in the result.

We then show that neural networks can be trained to predict both the exchange coupling interaction and the distance distribution in parallel. They were shown to provide a confident estimate at the magnitude of the exchange interaction while offering varying success when estimating the distance distribution.

Text
steven_worswick_mphil_compchem_september2020 - Version of Record
Download (5MB)
Text
Permission to deposit thesis - signed
Restricted to Repository staff only

More information

Published date: 2020

Identifiers

Local EPrints ID: 453032
URI: http://eprints.soton.ac.uk/id/eprint/453032
PURE UUID: b243f5fd-8905-4877-8913-1975f38aca02
ORCID for Steven Graham Worswick: ORCID iD orcid.org/0000-0002-5864-9684
ORCID for Ilya Kuprov: ORCID iD orcid.org/0000-0003-0430-2682

Catalogue record

Date deposited: 07 Jan 2022 17:41
Last modified: 17 Mar 2024 03:28

Export record

Contributors

Author: Steven Graham Worswick ORCID iD
Thesis advisor: Ilya Kuprov ORCID iD

Download statistics

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Atom RSS 1.0 RSS 2.0

Contact ePrints Soton: eprints@soton.ac.uk

ePrints Soton supports OAI 2.0 with a base URL of http://eprints.soton.ac.uk/cgi/oai2

This repository has been built using EPrints software, developed at the University of Southampton, but available to everyone to use.

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×