Skip to Main content Skip to Navigation
Journal articles

Two Classes Linear Discrimination, a Min-Max Approach

Abstract : In this paper we propose a new dichotomic linear discrimination algorithm based on a criterion recursively defined from the min and max functions. This criterion allows us, using a gradient algorithm, to find two hyperplanes which split two classes in an optimal way. We give two examples of the possibilities of the method.
Document type :
Journal articles
Complete list of metadatas

Cited literature [4 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-00340822
Contributor : Philippe Castagliola <>
Submitted on : Tuesday, May 19, 2020 - 5:28:20 PM
Last modification on : Wednesday, June 24, 2020 - 4:19:43 PM

File

1989_PRL.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00340822, version 1

Citation

Philippe Castagliola, Bernard Dubuisson. Two Classes Linear Discrimination, a Min-Max Approach. Pattern Recognition Letters, Elsevier, 1989, 10, pp.281-287. ⟨hal-00340822⟩

Share

Metrics

Record views

169

Files downloads

30