Krein Space Methods for Structured Data

Learning from structured data, including sequences and graphs, is a significant and central challenge in machine learning that has far-reaching applications across numerous disciplines, including chemistry and biochemistry. Kernel methods are undoubtedly an essential tool in this challenge. Their us...

Full description

Bibliographic Details
Main Author: Redshaw, Joseph
Format: Thesis (University of Nottingham only)
Language:English
Published: 2024
Subjects:
Online Access:https://eprints.nottingham.ac.uk/78390/
_version_ 1848801077181284352
author Redshaw, Joseph
author_facet Redshaw, Joseph
author_sort Redshaw, Joseph
building Nottingham Research Data Repository
collection Online Access
description Learning from structured data, including sequences and graphs, is a significant and central challenge in machine learning that has far-reaching applications across numerous disciplines, including chemistry and biochemistry. Kernel methods are undoubtedly an essential tool in this challenge. Their use of a similarity function, known as a kernel function, facilitates learning complex relationships from data of arbitrary structure. However, many expressive notions of similarity are not valid kernel functions, meaning they are not applicable to standard kernel methods. Krein space methods are a potential solution to this problem, as they generalise kernel methods to a much larger class of similarity functions. In this thesis, we explore the application of Krein space methods to structured data. Focusing on problems in chemistry and biochemistry, in which structured data and domain-specific similarity measures are commonplace, we investigate to what extent Krein space methods can be utilised to develop supervised learning models for structured data. In particular, we develop models to identify translation initiation sites in nucleic acid sequences, predict the yield of a carbon-nitrogen cross coupling reaction and identify peptides exhibiting antimicrobial properties. We find that the resulting performance of the models is highly dependent on the choice of similarity function and that Krein space methods outperform standard kernel methods in some, but not all, of the domains considered.
first_indexed 2025-11-14T21:01:43Z
format Thesis (University of Nottingham only)
id nottingham-78390
institution University of Nottingham Malaysia Campus
institution_category Local University
language English
last_indexed 2025-11-14T21:01:43Z
publishDate 2024
recordtype eprints
repository_type Digital Repository
spelling nottingham-783902024-07-24T04:44:17Z https://eprints.nottingham.ac.uk/78390/ Krein Space Methods for Structured Data Redshaw, Joseph Learning from structured data, including sequences and graphs, is a significant and central challenge in machine learning that has far-reaching applications across numerous disciplines, including chemistry and biochemistry. Kernel methods are undoubtedly an essential tool in this challenge. Their use of a similarity function, known as a kernel function, facilitates learning complex relationships from data of arbitrary structure. However, many expressive notions of similarity are not valid kernel functions, meaning they are not applicable to standard kernel methods. Krein space methods are a potential solution to this problem, as they generalise kernel methods to a much larger class of similarity functions. In this thesis, we explore the application of Krein space methods to structured data. Focusing on problems in chemistry and biochemistry, in which structured data and domain-specific similarity measures are commonplace, we investigate to what extent Krein space methods can be utilised to develop supervised learning models for structured data. In particular, we develop models to identify translation initiation sites in nucleic acid sequences, predict the yield of a carbon-nitrogen cross coupling reaction and identify peptides exhibiting antimicrobial properties. We find that the resulting performance of the models is highly dependent on the choice of similarity function and that Krein space methods outperform standard kernel methods in some, but not all, of the domains considered. 2024-07-24 Thesis (University of Nottingham only) NonPeerReviewed application/pdf en cc_by https://eprints.nottingham.ac.uk/78390/1/JR_Thesis.pdf Redshaw, Joseph (2024) Krein Space Methods for Structured Data. PhD thesis, University of Nottingham. structured data kernel methods Kreın space methods antimicrobial peptides
spellingShingle structured data
kernel methods
Kreın space methods
antimicrobial peptides
Redshaw, Joseph
Krein Space Methods for Structured Data
title Krein Space Methods for Structured Data
title_full Krein Space Methods for Structured Data
title_fullStr Krein Space Methods for Structured Data
title_full_unstemmed Krein Space Methods for Structured Data
title_short Krein Space Methods for Structured Data
title_sort krein space methods for structured data
topic structured data
kernel methods
Kreın space methods
antimicrobial peptides
url https://eprints.nottingham.ac.uk/78390/