Abstract:Dimensionality reduction is a key preprocessing step for many applications. Until our knowledge, unsupervised approaches such as PCA and ICA do not take label information of the original data into account, so a supervised approach such as Linear discriminant analysis (LDA) performs better on many classification tasks. Unfortunately, the classical LDA approach has shortcomings, such as the well-known small size problem, the heteroscedastic problem and the (C-1) low rank problem. The (C-1) low rank problem greatly limits the dimension of the extracted features. In addition, the calculation of the between-class and within-class scatter matrices in the classical LDA approach actually only takes account of the Mahalanobis distance like covariance distance of data centers and each data class, so if the dataset has very few classes or the data distribution of each class is not Gaussian-like but has some spatial structure in the feature space instead, classical LDA does not work well. In this paper we propose a dimensionality reduction approach which avoids the limitations of classical LDA and improves handling of the between-class scatter matrix. Our approach approach takes the distribution of data in each class into consideration to calculate the projection matrix. It does not assume that the data distribution of each class approximates Gaussian; each can have its own spatial structure. Experiments show that our method can obtain better projection directions than the classical LDA approach and greatly improve the classification accuracy. In addition, our approach is able to reconstruct the original signal well, while the classical LDA approach ignores the reconstruction property.

Multimodal Linear Discriminant Analysis Via Structural Sparsity

Adaptive Local Linear Discriminant Analysis

Generalized robust linear discriminant analysis for jointly sparse learning

Linear Discriminant Regularized Regression

Understanding and Evaluating Sparse Linear Discriminant Analysis

Dimension reduction of multimodal data by auto-weighted local discriminant analysis

Semi-Supervised Multi-Label Linear Discriminant Analysis

Learning in Multimodal and Mixmodal Data: Locality Preserving Discriminant Analysis with Kernel and Sparse Representation Techniques.

Sr-LDA:Sparse and Reduced-Rank Linear Discriminant Analysis for High Dimensional Matrix

A procedure of linear discrimination analysis with detected sparsity structure for high-dimensional multi-class classification

Multi-label Linear Discriminant Analysis with Locality Consistency.

Regularized Deep Linear Discriminant Analysis

Robust Linear Discriminant Analysis Using Ratio Minimization of L1,2-Norms

Linear discriminant analysis with spectral regularization

Reconstructable generalized maximum scatter difference discriminant analysis

Robust sparse manifold discriminant analysis

Sparse linear discriminant analysis by thresholding for high dimensional data

Linear Sequence Discriminant Analysis: A Model-Based Dimensionality Reduction Method for Vector Sequences

Neighborhood linear discriminant analysis

Minimally Informed Linear Discriminant Analysis: training an LDA model with unlabelled data

Two-Stage Regularized Linear Discriminant Analysis for 2-D Data.