Show simple item record  

dc.contributor.advisorPfahringer, Bernhard
dc.contributor.advisorFrank, Eibe
dc.contributor.authorSahito, Attaullah
dc.date.accessioned2021-12-02T23:39:12Z
dc.date.available2021-12-02T23:39:12Z
dc.date.issued2021
dc.identifier.urihttps://hdl.handle.net/10289/14678
dc.description.abstractArtificial neural networks achieve state-of-the-art performance when trained on a vast number of labelled examples. Still, they can easily overfit training examples when few labelled examples are available. The requirement to have labels for all training examples is a strong limitation of standard supervised machine learning. This can be addressed by applying semi-supervised learning methods that extend supervised learning and use unlabelled examples. Self-training is the most basic and generic semi-supervised approach. In self-training, a model is trained iteratively on both labelled and pseudo-labelled examples obtained from previous iterations. This thesis focuses on the task of investigating different variants of self-training by applying metric learning, transfer learning, and self-supervised learning. The first part of this thesis investigates how metric learning can be applied to self-training. This is achieved by applying several metric learning losses for the training of feedforward neural networks. Experimental results show that triplet loss – a metric learning loss – can achieve better results than cross-entropy loss with simple neural networks. For improving the performance of self-training, the second part of the thesis investigates applying large neural networks and pretraining on various image sizes of ImageNet with different loss functions. Experimental results show that pretraining always improves the predictive performance of the model. Pretraining on smaller image sizes with cross-entropy loss provides the highest performance. In the third part of this thesis, several self-training methods are developed using self-supervised learning. Geometric transformation-based self-supervised learning is applied to unlabelled examples. The experimental results indicate that applying self-supervised learning for only the first iteration achieves better performance than using it in all iterations of self-training.
dc.format.mimetypeapplication/pdf
dc.language.isoen
dc.publisherThe University of Waikato
dc.rightsAll items in Research Commons are provided for private study and research purposes and are protected by copyright with all rights reserved unless otherwise indicated.
dc.subjectSelf-training
dc.subjectSemi-supervised Learning
dc.subjectMetric Learning
dc.subjectTransfer Learning
dc.subjectSelf-supervised Learning
dc.subject.lcshNeural networks (Computer science)
dc.subject.lcshMachine learning
dc.subject.lcshArtificial intelligence
dc.subject.lcshComputer vision
dc.subject.lcshImage processing -- Classification
dc.subject.lcshPattern recognition systems
dc.titleA study of self-training variants for semi-supervised image classification
dc.typeThesis
thesis.degree.grantorThe University of Waikato
thesis.degree.levelDoctoral
thesis.degree.nameDoctor of Philosophy (PhD)
dc.date.updated2021-11-25T04:10:36Z
pubs.place-of-publicationHamilton, New Zealanden_NZ


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record