Previous |  Up |  Next

Article

Title: Construction of nonlinear discrimination function based on the MDL criterion (English)
Author: Sato, Manabu
Author: Kudo, Mineichi
Author: Toyama, Jun
Author: Shimbo, Masaru
Language: English
Journal: Kybernetika
ISSN: 0023-5954
Volume: 34
Issue: 4
Year: 1998
Pages: [467]-472
Summary lang: English
.
Category: math
.
Summary: Although a nonlinear discrimination function may be superior to linear or quadratic classifiers, it is difficult to construct such a function. In this paper, we propose a method to construct a nonlinear discrimination function using Legendre polynomials. The selection of an optimal set of Legendre polynomials is determined by the MDL (Minimum Description Length) criterion. Results using many real data show the effectiveness of this method. (English)
Keyword: nonlinear discrimination function
Keyword: Legendre polynomials
MSC: 62H25
MSC: 62H30
MSC: 68T05
idZBL: Zbl 1274.68348
.
Date available: 2009-09-24T19:19:20Z
Last updated: 2015-03-28
Stable URL: http://hdl.handle.net/10338.dmlcz/135233
.
Reference: [2] Murphy P. M., Aha D. W.: UCI Repository of Machine Learning Databases [Machine–Readable Data Repository].University of California, Department of Information and Computer Science, Irvine 1991
Reference: [3] Rissanen J.: A universal prior for integers and estimation by minimum description length.Ann. Statist. 11 (1983), 416–431 Zbl 0513.62005, MR 0696056, 10.1214/aos/1176346150
Reference: [4] Stone M.: Cross–Validatory choice and assessment of statistical predictions.J. Roy. Statist. Soc. 36 (1974), 111–147 Zbl 0308.62063, MR 0356377
.

Files

Files Size Format View
Kybernetika_34-1998-4_19.pdf 709.2Kb application/pdf View/Open
Back to standard record
Partner of
EuDML logo