Show simple item record

dc.contributor.supervisor Fowler, Kent (Anthropology) en_US
dc.contributor.author Koval, Oksana
dc.date.accessioned 2018-05-04T20:06:32Z
dc.date.available 2018-05-04T20:06:32Z
dc.date.issued 2018
dc.date.submitted 2018-05-04T19:46:38Z en
dc.identifier.uri http://hdl.handle.net/1993/33021
dc.description.abstract Objective: 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.subject machine learning en_US
dc.subject supervised en_US
dc.subject supervised machine learning en_US
dc.subject classification en_US
dc.subject nearest neighbour en_US
dc.subject nearest neighbor en_US
dc.subject k nearest neighbour en_US
dc.subject archaeology en_US
dc.subject wndchrm en_US
dc.subject WND-CHRM en_US
dc.subject wndcharm en_US
dc.subject wnd en_US
dc.subject weighted nearest neighbour en_US
dc.subject weighted nearest neighbor en_US
dc.subject weighted nearest neighbour distance en_US
dc.subject weighted nearest distance en_US
dc.subject weighted neighbour distance using compound hierarchy of algorithms representing morphology en_US
dc.subject algorithm en_US
dc.subject pattern recognition en_US
dc.subject ceramic fabric analysis en_US
dc.subject ceramic en_US
dc.subject ceramic fabric en_US
dc.subject ceramics en_US
dc.subject pot en_US
dc.subject pottery en_US
dc.subject ceramic fabric en_US
dc.subject pottery fabric en_US
dc.subject paste en_US
dc.subject ceramic paste en_US
dc.subject pottery paste en_US
dc.subject fabric analysis en_US
dc.subject shaping en_US
dc.subject ceramic shaping en_US
dc.subject pottery shaping en_US
dc.subject vessel shaping en_US
dc.subject identification en_US
dc.subject data collection en_US
dc.subject image aquisition en_US
dc.subject dinoxcope en_US
dc.subject digital microscope en_US
dc.subject image processing en_US
dc.subject tile en_US
dc.subject tiling en_US
dc.subject ceramic classification en_US
dc.subject pottery classification en_US
dc.subject computer ceramic classification en_US
dc.subject computer analysis ceramics en_US
dc.subject pottery computer analysis en_US
dc.subject ceramic classification en_US
dc.subject dataset en_US
dc.subject datasets en_US
dc.subject pottery dataset en_US
dc.subject ceramic dataset en_US
dc.subject fabric dataset en_US
dc.subject ceramic fabric dataset en_US
dc.subject ceramic paste dataset en_US
dc.subject shaping methods en_US
dc.subject shaping technique en_US
dc.subject zulu en_US
dc.subject south africa en_US
dc.subject ethnoarchaeology en_US
dc.subject experimental en_US
dc.subject superrvised en_US
dc.subject fabric dataset en_US
dc.subject training set en_US
dc.subject training dataset en_US
dc.subject guidelines en_US
dc.subject computer analysis en_US
dc.subject new approaches en_US
dc.subject method en_US
dc.subject methodology en_US
dc.subject anthropology en_US
dc.subject programming en_US
dc.subject dataset guideline en_US
dc.subject ceramic production analysis en_US
dc.subject technological style en_US
dc.subject ceramic technological style en_US
dc.subject identify individual potter en_US
dc.subject individual en_US
dc.subject identify en_US
dc.subject provenience en_US
dc.subject homogeneity en_US
dc.subject accuracy en_US
dc.subject dataset design en_US
dc.subject ceramic dataset design en_US
dc.subject ceramic images dataset en_US
dc.subject ceramic images training set en_US
dc.subject ceramic fabric training set en_US
dc.subject ceramic fabric dataset en_US
dc.subject machine learning in archaeology en_US
dc.subject machine learning in ceramic analysis en_US
dc.subject image analysis en_US
dc.title Using supervised machine learning to classify ceramic fabrics en_US
dc.degree.discipline Anthropology en_US
dc.contributor.examiningcommittee Hoppa, Robert (Anthropology) en_US
dc.contributor.examiningcommittee Lawall, Mark (Classics) en_US
dc.degree.level Master of Arts (M.A.) en_US
dc.description.note May 2018 en_US


Files in this item

This item appears in the following Collection(s)

Show simple item record

View Statistics