Abstract:Linear discriminant analysis (LDA) represents a simple yet powerful technique for partitioning a p-dimensional feature vector into one of K classes based on a linear projection learned from N labeled observations. However, it is well-established that in the high-dimensional setting (p > N) the underlying projection estimator degenerates. Moreover, any linear discriminate function involving a large number of features may be difficult to interpret. To ameliorate these issues, two general categories of sparse LDA modifications have been proposed, both to reduce the number of active features and to stabilize the resulting projections. The first, based on optimal scoring, is more straightforward to implement and analyze but has been heavily criticized for its ambiguous connection with the original LDA formulation. In contrast, a second strategy applies sparse penalty functions directly to the original LDA objective but requires additional heuristic trade-off parameters, has unknown global and local minima properties, and requires a greedy sequential optimization procedure. In all cases the choice of sparse regularizer can be important, but no rigorous guidelines have been provided regarding which penalty might be preferable. Against this backdrop, we winnow down the broad space of candidate sparse LDA algorithms and promote a specific selection based on optimal scoring coupled with a particular, complementary sparse regularizer. This overall process ultimately progresses our understanding of sparse LDA in general, while leading to targeted modifications of existing algorithms that produce superior results in practice on three high-dimensional gene data sets.

Least squares online linear discriminant analysis

Theoretical Analysis of Direct LDA in Small Sample Size Problem

Least Square Incremental Linear Discriminant Analysis

A Local Online Learning Approach for Non-linear Data.

Sequential Linear Discriminant Analysis On Undersampled Problems

Adaptive Local Linear Discriminant Analysis

Training Linear Discriminant Analysis in Linear Time

Incremental Regularized Least Squares for Dimensionality Reduction of Large-Scale Data.

An Incremental Linear Discriminant Analysis Using Fixed Point Method

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

Understanding and Evaluating Sparse Linear Discriminant Analysis

A Flexible and Efficient Algorithm for Regularized Fisher Discriminant Analysis.

Efficient and Robust Sparse Linear Discriminant Analysis for Data Classification

From NLDA to LDA/GSVD: a Modified NLDA Algorithm

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

Linear Discriminant Regularized Regression

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

Making Fisher Discriminant Analysis Scalable.

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

Varying Coefficient Linear Discriminant Analysis for Dynamic Data

On the Equivalence of Linear Discriminant Analysis and Least Squares Regression