A generalized optimal set of discriminant vectors

Pattern RecognitionPublished 1 July 1992

Ke Liu, Yong-Qing Cheng, Jing-Yu Yang

DOI10.1016/0031-3203(92)90136-7

Source URL

Citations101

SJR quartileQ1

SJR score2.06

SNIP2.67

TL;DR

The experimental results show that the present method is superior to the Foley-Sammon method, the positive pseudoinverse method, and the matrix rank decomposition method in terms of correct classification rate.

Abstract

A generalized optimal set of discriminant vectors for linear feature extraction is presented. First, the criteria of selecting the generalized optimal discriminant vectors are introduced, and then a unified solving method is derived to solve the vectors of the generalized optimal set in both cases of a large number of samples and a small number of samples. The experimental results show that the present method is superior to the Foley-Sammon method (Foley and Sammon, IEEE Trans. Comput.24, 281–289 (1975)), the positive pseudoinverse method (Tian et al., Opt. Engng25(7), 834–839 (1986)), the perturbation method (Hong and Yang, Pattern Recognition24, 317–324 (1991)), and the matrix rank decomposition method (Cheng et al., Pattern Recognition25, 101–111 (1992)) in terms of correct classification rate.

Keywords

Computer ScienceEngineering

Annals of EugenicsTHE USE OF MULTIPLE MEASUREMENTS IN TAXONOMIC PROBLEMS

14,727 Citations1936Ronald Aylmer Fisher

View PDF

The algebraic eigenvalue problem

5,785 Citations1965J. H. Wilkinson

IEEE Transactions on ComputersAn Optimal Set of Discriminant Vectors

404 Citations1975Donald H. Foley, John W. Sammon

A new method for the extraction of features in a two-class pattern recognition problem is derived that is based entirely upon discrimination or separability as opposed to the more common approach of fitting.

Pattern RecognitionOptimal discriminant plane for a small number of samples and design method of classifier on the plane

302 Citations1991Zi-Quan Hong, Jing-Yu Yang

It is proved that the generalized eigenequations on the optimal discriminant plane are stable in respect of eigenvalues and the generalized Eigenvectors are indeed the optimal discriminate directions, if the perturbation is subject to some certain conditions.

Pattern RecognitionAlgebraic feature extraction of image for recognition

281 Citations1991Zi-Quan Hong

This paper proves that SV feature vector has some important properties of algebraic and geometric invariance, and insensitiveness to noise, and these properties are very useful for the description and recognition of images.

IEEE Transactions on ComputersAn Optimal Discriminant Plane

118 Citations1970John W. Sammon

A piecewise linear method is described which is being used in the on-line pattern analysis and recognition system (OLPARS) and it is shown that it can be used to discriminate between vector samples from two classes.

Optical EngineeringImage Classification By The Foley-Sammon Transform

86 Citations1986Qing Tian, M. Barbero +2 more

A set of computer simulation results shows that the Foley-Sammon transform provides good classification performance and that a proper number of training images is essential to ensure a low error rate of classification.

Pattern RecognitionOptimal fisher discriminant analysis using the rank decomposition

64 Citations1992Yong-Qing Cheng, Yong-Ming Zhuang +1 more

In this paper, the conception of the rank decomposition of matrices is first introduced, and a new method for calculating the Fisher optimal discriminant vector is presented which is particularly well suited to the case of a small number of samples in the sense that the scatter matrices are rank-deficient.

Journal of the Optical Society of America AComparison of statistical pattern-recognition algorithms for hybrid processing II Eigenvector-based algorithm

39 Citations1988Qinglin Tian, Y. Fainman +1 more

It is shown that all eigen vector-based algorithms can be represented in a generalized eigenvector form, and the classification performance with discriminant functions of the F-S, HTC, and F-K with the LDF and GMF algorithms and the linear-mapping- based algorithms with the eigen vectors.

Journal of the Optical Society of America AComparison of statistical pattern-recognition algorithms for hybrid processing I Linear-mapping algorithms

26 Citations1988Qinglin Tian, Y. Fainman +2 more

The relations among various linear-mapping-based algorithms are studied by formulating a more general unified pseudoinverse algorithm and it is shown that the least-squares linear-Mapping technique, the simplified least-Squares linear, the synthetic discriminant function, the equal-correlation-peak method, and the Caulfield–Maloney filter are in fact all special cases of the unified pseudo inverse algorithm.