South Dakota State University
Classification Error of Gaussian and Transformed Gaussian Variates
Abstract
dc:description.abstractAs indicated by the title and by the large proportion of pages devoted to the subject, transformation of variables is the primary topic of this thesis. The transformation of partially overlapping, correlated, two-dimensional, Gaussian distributed random variates to partially overlapping, independent, one-dimensional, Cauchy distributed random variates involves a reduction in dimensionality by one. Consequently, the equation defining the receiver is simplified, and the expressions for predicting the classification error are easier to formulate. The equation required to effect this transformation is z=Y1/Y2 , where Y1 , Y 2 are the standardized correlated Gaussian variates and z is the independent Cauchy rand om variate.
Degree
thesis:*- Name thesis:degree_name
- Master of Science (MS)
- Level thesis:degree_level
- Thesis - Open Access
- Discipline thesis:degree_discipline
- Electrical Engineering
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Boyd, John E.
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://openprairie.sdstate.edu/etd/3690