Coupling Recursive Hyperspheric Classification with Linear Discriminant Analysis for Improved Results

Author

Reed, Salyer B. ; Reed, Tyson R. C. ; Dascalu, Sergiu M.

Author_Institution

Dept. of Comput. Sci. & Eng., Univ. of Nevada, Reno, NV, USA

fYear

2013

fDate

15-17 April 2013

Firstpage

596

Lastpage

601

Abstract

Recursive Hyper spheric Classification (RHC) can accurately and succinctly classify large datasets by dissecting labeled vectors into their constituent groups, or hyper spheres. While RHC is a powerful classification tool, coupling RHC with other linear classifiers enhances the ability and accuracy of the classification system, improving recognition of unlabeled vectors. In this paper, RHC is paired with Linear Discriminant Analysis (LDA) for improved classification and recognition rates.

Keywords

learning (artificial intelligence); pattern classification; vectors; LDA; classification rates; classification system; coupling RHC; hyper spheres; linear classifiers; linear discriminant analysis; powerful classification tool; recognition rates; recursive hyperspheric classification; unlabeled vectors recognition; Classification algorithms; Couplings; Euclidean distance; Linear discriminant analysis; Support vector machine classification; Vectors; Dimensional Reduction; Linear Discriminant Analysis; Machine Learning; Recursive Hyperspheric Classification; Wine Dataset;

fLanguage

English

Publisher

ieee

Conference_Titel

Information Technology: New Generations (ITNG), 2013 Tenth International Conference on

Conference_Location

Las Vegas, NV

Print_ISBN

978-0-7695-4967-5

Type

conf

DOI

10.1109/ITNG.2013.91

Filename

6614371