Classification in high dimensional feature spaces / by H.O. van Dyk

Boloka/Manakin Repository

Show simple item record

dc.contributor.author Van Dyk, Hendrik Oostewald
dc.date.accessioned 2011-04-06T14:12:49Z
dc.date.available 2011-04-06T14:12:49Z
dc.date.issued 2009
dc.identifier.uri http://hdl.handle.net/10394/4091
dc.description Thesis (M.Ing. (Computer Engineering))--North-West University, Potchefstroom Campus, 2009.
dc.description.abstract In this dissertation we developed theoretical models to analyse Gaussian and multinomial distributions. The analysis is focused on classification in high dimensional feature spaces and provides a basis for dealing with issues such as data sparsity and feature selection (for Gaussian and multinomial distributions, two frequently used models for high dimensional applications). A Naïve Bayesian philosophy is followed to deal with issues associated with the curse of dimensionality. The core treatment on Gaussian and multinomial models consists of finding analytical expressions for classification error performances. Exact analytical expressions were found for calculating error rates of binary class systems with Gaussian features of arbitrary dimensionality and using any type of quadratic decision boundary (except for degenerate paraboloidal boundaries). Similarly, computationally inexpensive (and approximate) analytical error rate expressions were derived for classifiers with multinomial models. Additional issues with regards to the curse of dimensionality that are specific to multinomial models (feature sparsity) were dealt with and tested on a text-based language identification problem for all eleven official languages of South Africa.
dc.publisher North-West University
dc.subject Naïve Bayesian en
dc.subject Maximum likelihood en
dc.subject Curse of dimensionality en
dc.subject Gaussian distribution en
dc.subject Multinomial distribution en
dc.subject Feature selection en
dc.subject Data sparsity en
dc.subject Chi-square variates en
dc.subject Hyperboloidal decision boundaries en
dc.title Classification in high dimensional feature spaces / by H.O. van Dyk en
dc.type Thesis en
dc.description.thesistype Masters

Files in this item

This item appears in the following Collection(s)

  • ETD@PUK [6252]
    This collection contains the original digitized versions of research conducted at the North-West University (Potchefstroom Campus)

Show simple item record

Search the NWU Repository

Advanced Search


My Account