Abstract:Linear Discriminant Analysis (LDA) is a well-known method for feature extraction and dimension reduction. It has been used widely in many applications involving high-dimensional data, such as image and text classification. An intrinsic limitation of classical LDA is the so-called singularity problems; that is, it fails when all scatter matrices are singular. Many LDA extensions were proposed in the past to overcome the singularity problems. Among these extensions, PCA+LDA, a two-stage method, received relatively more attention. In PCA+LDA, the LDA stage is preceded by an intermediate dimension reduction stage using Principal Component Analysis (PCA). Most previous LDA extensions are computationally expensive, and not scalable, due to the use of Singular Value Decomposition or Generalized Singular Value Decomposition. In this paper, we propose a two-stage LDA method, namely LDA/QR, which aims to overcome the singularity problems of classical LDA, while achieving efficiency and scalability simultaneously. The key difference between LDA/QR and PCA+LDA lies in the first stage, where LDA/QR applies QR decomposition to a small matrix involving the class centroids, while PCA+LDA applies PCA to the total scatter matrix involving all training data points. We further justify the proposed algorithm by showing the relationship among LDA/QR and previous LDA methods. Extensive experiments on face images and text documents are presented to show the effectiveness of the proposed algorithm.

Linear Extension of Graph Embedding Based on QR Decomposition and Face Recognition

Dlda/Qra Robust Direct Lda Algorithm For Face Recognition And Its Theoretical Foundation

Feature extraction based on graph discriminant embedding and its applications to face recognition

Graph Embedding Algorithms Based on Neighborhood Discriminant Embedding for Face Recognition

Face Recognition Base On Uncorrelated Linear Extension Of Graph Embedding

Local Graph Embedding Based on Maximum Margin Criterion (LGE/MMC) for Face Recognition.

Face Feature Extraction and Recognition Via Local Binary Pattern and Two-Dimensional Locality Preserving Projection

Face Linear Discriminant Analysis Based on Graph Embedding and Regularization

Optimal Discriminant Analysis Based on Kernel Extension of Graph Embedding and Face Recognition

Graph Embedding Discriminant Analysis for Face Recognition

Sparse Graph Embedding Based On The Fuzzy Set For Image Classification

A novel graph-embedding feature extraction method based on local and global inter-class graph

Label Information-Based Weighted Regularized Sparsity Preserving Embedding for Face Recognition

Extended Locally Linear Embedding with Gabor Wavelets for Face Recognition.

A Two-Stage Linear Discriminant Analysis Via QR-decomposition.

Uncorrelated and Discriminative Graph Embedding for Face Recognition

MiLDA: A Graph Embedding Approach to Multi-View Face Recognition

An Efficient Regularized Neighborhood Discriminant Analysis Through QR Decomposition

Local Graph Embedding Discriminant Analysis for Face Recognition with Single Training Sample Per Person

Two-dimensional Local Graph Embedding Discriminant Analysis (2DLGEDA) with Its Application to Face and Palm Biometrics

Bilinear Neighborhood Discriminant Embedding And Its Non-Iterative Solution Algorithm