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.

A Novel Dimensionality Reduction Method with Discriminative Generalized Eigen-Decomposition.

Computational and Theoretical Analysis of Supervised Dimensionality Reduction

A Discriminant Graph Nonnegative Matrix Factorization Approach to Computer Vision

Discriminative Unsupervised Dimensionality Reduction.

A novel dimensionality reduction method: Similarity order preserving discriminant analysis

Supervised Dimensionality Reduction Technology of Generalized Discriminant Component Analysis and Its Kernelization Forms

An Incremental Dimensionality Reduction Method on Discriminant Information for Pattern Classification.

Transferred Dimensionality Reduction

Optimal Dimensionality Discriminant Analysis and Its Application to Image Recognition

Nonparametric Discriminant Multi-Manifold Learning for Dimensionality Reduction

Fusion of effective dimension reduction and discriminative dictionary learning for high-dimensional classification

A Novel Discriminant Criterion with Application to Face Recognition

An Orthogonal Locality and Globality Dimensionality Reduction Method Based on Twin Eigen Decomposition

Towards the Optimal Discriminant Subspace

GMFLLM: A General Manifold Framework Unifying Three Classic Models for Dimensionality Reduction

Deep Dimension Reduction for Supervised Representation Learning

Discriminative Unsupervised 2D Dimensionality Reduction with Graph Embedding

Deep Nonlinear Sufficient Dimension Reduction

C2DNDA: A Deep Framework for Nonlinear Dimensionality Reduction

Reconstructable generalized maximum scatter difference discriminant analysis

Adaptive and Fuzzy Locality Discriminant Analysis for Dimensionality Reduction