A Direct Estimation Approach to Sparse Linear Discriminant Analysis

TL;DR

This paper introduces a new sparse linear discriminant analysis method that estimates the product of the precision matrix and mean difference directly, offering computational efficiency and robustness in high-dimensional classification tasks.

Contribution

It proposes the LPD rule, a direct estimation approach using linear programming, which improves over existing methods by handling cases with high sparsity and computational efficiency.

Findings

LPD rule performs well even when $\\O$ and/or $\de$ are not estimable.

The method has strong theoretical guarantees including consistency and convergence.

Empirical results show superior performance on real high-dimensional datasets.

Abstract

This paper considers sparse linear discriminant analysis of high-dimensional data. In contrast to the existing methods which are based on separate estimation of the precision matrix $\O$ and the difference $\de$ of the mean vectors, we introduce a simple and effective classifier by estimating the product $\O\de$ directly through constrained $\ell_1$ minimization. The estimator can be implemented efficiently using linear programming and the resulting classifier is called the linear programming discriminant (LPD) rule. The LPD rule is shown to have desirable theoretical and numerical properties. It exploits the approximate sparsity of $\O\de$ and as a consequence allows cases where it can still perform well even when $\O$ and/or $\de$ cannot be estimated consistently. Asymptotic properties of the LPD rule are investigated and consistency and rate of convergence results are given. The LPD classifier has superior finite sample performance and significant computational advantages over the existing methods that require separate estimation of $\O$ and $\de$. The LPD rule is also applied to analyze real datasets from lung cancer and leukemia studies. The classifier performs favorably in comparison to existing methods.