dc.contributor.author | Sugg, B | |
dc.date.accessioned | 2018-12-11T18:46:26Z | |
dc.date.issued | 2018-12-10 | |
dc.description.abstract | The Convolutional Neural Network (CNN) is a machine learning model which excels in tasks that exhibit spatially local correlation of features, for example, image classification. However, as a model, it is susceptible to the issues caused by local minima, largely due to the fully-connected neural network which is typically used in the final layers for classification. This work investi- gates the effect of replacing the fully-connected neural network with a Support Vector Machine (SVM). It names the resulting model the Convolutional Support Vector Machine (CSVM) and proposes two methods for training. The first method uses a linear SVM and it is described in the primal. The second method can be used to learn a SVM with a non-linear kernel by casting the optimisation as a Multiple Kernel Learning problem. Both methods learn the convolutional filter weights in conjunction with the SVM parameters. The linear CSVM (L-CSVM) and kernelised CSVM (K-CSVM) in this work each use a single convolutional filter, however, approaches are described which may be used to extend the K-CSVM with multiple filters per layer and with multiple convolutional layers. The L-CSVM and K-CSVM show promising results on the MNIST and CIFAR-10 benchmark datasets. | en_GB |
dc.identifier.uri | http://hdl.handle.net/10871/35102 | |
dc.language.iso | en | en_GB |
dc.publisher | University of Exeter | en_GB |
dc.subject | Image Classification | en_GB |
dc.subject | SVM | en_GB |
dc.subject | Convolution | en_GB |
dc.title | Convolutional Support Vector Machines For Image Classification | en_GB |
dc.type | Thesis or dissertation | en_GB |
dc.date.available | 2018-12-11T18:46:26Z | |
dc.contributor.advisor | Everson, R | en_GB |
dc.contributor.advisor | Fieldsend, J | en_GB |
dc.publisher.department | College of Engineering, Mathematics and Physical Sciences | en_GB |
dc.rights.uri | http://www.rioxx.net/licenses/all-rights-reserved | en_GB |
dc.type.degreetitle | Masters by Research in Computer Science | en_GB |
dc.type.qualificationlevel | Masters | en_GB |
dc.type.qualificationname | MbyRes Dissertation | en_GB |
dcterms.dateAccepted | 2018-12-11 | |
rioxxterms.version | NA | en_GB |
rioxxterms.licenseref.startdate | 2018-11-22 | |
rioxxterms.type | Thesis | en_GB |
refterms.dateFOA | 2018-12-11T18:46:29Z | |