Using supervised machine learning to classify ceramic fabrics

Simple item page
dc.contributor.authorKoval, Oksana
dc.contributor.examiningcommitteeHoppa, Robert (Anthropology)en_US
dc.contributor.examiningcommitteeLawall, Mark (Classics)en_US
dc.contributor.supervisorFowler, Kent (Anthropology)en_US
dc.date.accessioned2018-05-04T20:06:32Z
dc.date.available2018-05-04T20:06:32Z
dc.date.issued2018
dc.date.submitted2018-05-04T19:46:38Zen
dc.degree.disciplineAnthropologyen_US
dc.degree.levelMaster of Arts (M.A.)en_US
dc.description.abstractObjective: The purpose of this thesis was to assess the ability of supervised machine learning to discriminate between images of ceramic fabrics and evaluate the requirements for creating an effective dataset. Method: Weighted Neighbour Distance using Compound Hierarchy of Algorithms Rep- resenting Morphology (wndchrm) algorithm, was applied to Zulu ceramic fabrics from South Africa. Wndchrm was used to extract thousands of image content descriptors, assign weights to the extracted features by learning their discriminative power from training examples, and classify unlabelled images by searching for a class with the nearest distance to the mean of the feature vector. Results: Wndchrm was successful in distinguishing ceramic fabrics by differences in paste. Comparable results were obtained in separate experiments designed to identify differences in paste by region, community and individual potters. In all cases, a sample size of 50 train- ing images per class was sufficient to produce 90% to 95% accuracy. In contrast, the experi- ments meant to identify fabrics by shaping techniques, represented by coiling and slab building, reached the accuracy of only 65% to 70%. Conclusion: The experiments show that wndchrm is an effective method for classification of ceramic fabrics, which can increase consistency in comparing and classifying ceramic fabrics by variation in paste. Creating and sharing training dataset libraries is the next step necessary for wide adoption of supervised machine learning in this field.en_US
dc.description.noteMay 2018en_US
dc.identifier.urihttp://hdl.handle.net/1993/33021
dc.language.isoengen_US
dc.rightsopen accessen_US
dc.subjectmachine learningen_US
dc.subjectsuperviseden_US
dc.subjectsupervised machine learningen_US
dc.subjectclassificationen_US
dc.subjectnearest neighbouren_US
dc.subjectnearest neighboren_US
dc.subjectk nearest neighbouren_US
dc.subjectarchaeologyen_US
dc.subjectwndchrmen_US
dc.subjectWND-CHRMen_US
dc.subjectwndcharmen_US
dc.subjectwnden_US
dc.subjectweighted nearest neighbouren_US
dc.subjectweighted nearest neighboren_US
dc.subjectweighted nearest neighbour distanceen_US
dc.subjectweighted nearest distanceen_US
dc.subjectweighted neighbour distance using compound hierarchy of algorithms representing morphologyen_US
dc.subjectalgorithmen_US
dc.subjectpattern recognitionen_US
dc.subjectceramic fabric analysisen_US
dc.subjectceramicen_US
dc.subjectceramic fabricen_US
dc.subjectceramicsen_US
dc.subjectpoten_US
dc.subjectpotteryen_US
dc.subjectceramic fabricen_US
dc.subjectpottery fabricen_US
dc.subjectpasteen_US
dc.subjectceramic pasteen_US
dc.subjectpottery pasteen_US
dc.subjectfabric analysisen_US
dc.subjectshapingen_US
dc.subjectceramic shapingen_US
dc.subjectpottery shapingen_US
dc.subjectvessel shapingen_US
dc.subjectidentificationen_US
dc.subjectdata collectionen_US
dc.subjectimage aquisitionen_US
dc.subjectdinoxcopeen_US
dc.subjectdigital microscopeen_US
dc.subjectimage processingen_US
dc.subjecttileen_US
dc.subjecttilingen_US
dc.subjectceramic classificationen_US
dc.subjectpottery classificationen_US
dc.subjectcomputer ceramic classificationen_US
dc.subjectcomputer analysis ceramicsen_US
dc.subjectpottery computer analysisen_US
dc.subjectceramic classificationen_US
dc.subjectdataseten_US
dc.subjectdatasetsen_US
dc.subjectpottery dataseten_US
dc.subjectceramic dataseten_US
dc.subjectfabric dataseten_US
dc.subjectceramic fabric dataseten_US
dc.subjectceramic paste dataseten_US
dc.subjectshaping methodsen_US
dc.subjectshaping techniqueen_US
dc.subjectzuluen_US
dc.subjectsouth africaen_US
dc.subjectethnoarchaeologyen_US
dc.subjectexperimentalen_US
dc.subjectsuperrviseden_US
dc.subjectfabric dataseten_US
dc.subjecttraining seten_US
dc.subjecttraining dataseten_US
dc.subjectguidelinesen_US
dc.subjectcomputer analysisen_US
dc.subjectnew approachesen_US
dc.subjectmethoden_US
dc.subjectmethodologyen_US
dc.subjectanthropologyen_US
dc.subjectprogrammingen_US
dc.subjectdataset guidelineen_US
dc.subjectceramic production analysisen_US
dc.subjecttechnological styleen_US
dc.subjectceramic technological styleen_US
dc.subjectidentify individual potteren_US
dc.subjectindividualen_US
dc.subjectidentifyen_US
dc.subjectprovenienceen_US
dc.subjecthomogeneityen_US
dc.subjectaccuracyen_US
dc.subjectdataset designen_US
dc.subjectceramic dataset designen_US
dc.subjectceramic images dataseten_US
dc.subjectceramic images training seten_US
dc.subjectceramic fabric training seten_US
dc.subjectceramic fabric dataseten_US
dc.subjectmachine learning in archaeologyen_US
dc.subjectmachine learning in ceramic analysisen_US
dc.subjectimage analysisen_US
dc.titleUsing supervised machine learning to classify ceramic fabricsen_US
dc.typemaster thesisen_US
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
koval_oksana.pdf
Size:
12.97 MB
Format:
Adobe Portable Document Format
Description:
Thesis
Download
License bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
license.txt
Size:
2.2 KB
Format:
Item-specific license agreed to upon submission
Description:
Download
Collections
FGS - Electronic Theses and Practica