Javascript must be enabled for the correct page display

Deep Support Vector Machines for Classification and Autoencoding Problems

Mark, R.W. van der (2013) Deep Support Vector Machines for Classification and Autoencoding Problems. Bachelor's Thesis, Artificial Intelligence.

Full text not available from this repository.

Abstract

Although the single-layer support vector machine is a popular machine learning method, previous research showed that the performance of the SVM for regression problems can be improved by adding a second layer of lower-level SVMs for feature extraction from the input patterns. In this thesis, a method for the application of the deep support vector machine for classification and autoencoding problems is presented. A two-layer support vector machine is compared to a regular SVM on eight classification datasets, and the results show that the DSVM outperforms the SVM on half of the tested datasets. For autoencoding problems, the DSVM is compared to a regular neural network autoencoder and a neural support vector machine. The results show that the DSVM also outperforms these methods.

Item Type: Thesis (Bachelor's Thesis)
Degree programme: Artificial Intelligence
Thesis type: Bachelor's Thesis
Language: English
Date Deposited: 15 Feb 2018 07:54
Last Modified: 15 Feb 2018 07:54
URI: https://fse.studenttheses.ub.rug.nl/id/eprint/11240

Actions (login required)

View Item View Item