Abstract:Linear discriminant analysis (LDA) has been a useful tool in pattern recognition and data analysis research and practice. While linearity of class boundaries cannot always be expected, nonlinear projections through pre-trained deep neural networks have served to map complex data onto feature spaces in which linear discrimination has served well. The solution to binary LDA is obtained by eigenvalue analysis of within-class and between-class scatter matrices. It is well known that the multiclass LDA is solved by an extension to the binary LDA, a generalised eigenvalue problem, from which the largest subspace that can be extracted is of dimension one lower than the number of classes in the given problem. In this paper, we show that, apart from the first of the discriminant directions, the generalised eigenanalysis solution to multiclass LDA does neither yield orthogonal discriminant directions nor maximise discrimination of projected data along them. Surprisingly, to the best of our knowledge, this has not been noted in decades of literature on LDA. To overcome this drawback, we present a derivation with a strict theoretical support for sequentially obtaining discriminant directions that are orthogonal to previously computed ones and maximise in each step the Fisher criterion. We show distributions of projections along these axes and demonstrate that discrimination of data projected onto these discriminant directions has optimal separation, which is much higher than those from the generalised eigenvectors of the multiclass LDA. Using a wide range of benchmark tasks, we present a comprehensive empirical demonstration that on a number of pattern recognition and classification problems, the optimal discriminant subspaces obtained by the proposed method, referred to as GO-LDA (Generalised Optimal LDA), can offer superior accuracy.

Linear discriminant analysis based on genetic algorithm

Optimal Feature Selection for Sparse Linear Discriminant Analysis and Its Applications in Gene Expression Data

Genetic generalized discriminant analysis and its applications

A novel weighted fuzzy LDA for face recognition using the genetic algorithm

GO-LDA: Generalised Optimal Linear Discriminant Analysis

A Novel Fisher Discriminant Approach Based On Genetic Algorithm

A Modified Algorithm for Generalized Discriminant Analysis

Understanding and Evaluating Sparse Linear Discriminant Analysis

MBLDA: A Novel Multiple Between-Class Linear Discriminant Analysis

Neighborhood linear discriminant analysis

From NLDA to LDA/GSVD: a Modified NLDA Algorithm

Largest Margin Neighbor Linear Discriminant Analysis

An Efficient Greedy Search Algorithm for High-Dimensional Linear Discriminant Analysis

Kernel-Based Fisher Minimum Discriminant Analysis and Face Recognition

Sparse L1-norm-based linear discriminant analysis

A Novel Method to Solve the Separation Problem of LDA.

Varying Coefficient Linear Discriminant Analysis for Dynamic Data

A Linear Discriminant Analysis Framework Based on Random Subspace for Face Recognition.

Feature extraction by dynamically revising the discriminant information in the objective space

Linear Laplacian Discrimination for Feature Extraction

New Linear Discriminant Analysis Based on Gaussian Mixture Model