Work

Matrix Linear Discriminant Analysis

Year: 2018

Type: preprint

Abstract: We propose a novel linear discriminant analysis approach for the classification of high-dimensional matrix-valued data that commonly arises from imaging studies. Motivated by the equivalence of the conventional linear discriminant analysis and the ordinary least squares, we consider an efficient nuclear norm penalized regression that encourages a low-rank structure. Theoretical properties including a non-asymptotic risk bound and a rank consistency result are established. Simulation studies and an application to electroencephalography data show the superior performance of the proposed method over the existing approaches.

Source: arXiv (Cornell University)

DOI: 10.48550/arxiv.1809.08746

Authors: Hu, Wei, Shen, Weining, Zhou, Hua, Kong, Dehan

Institutions: University of California, Irvine, University of Toronto, University of California, Los Angeles

Language: English

Topic: Statistical Methods and Inference

Subfield: Statistics and Probability

Field: Mathematics

Domain: Physical Sciences

Open Access status: green

Funders: National Science Foundation, National Institutes of Health, National Human Genome Research Institute

Awards: G410392181, Tensor Regressions and Applications in Neuroimaging Data Analysis, Collaborative Research: Studies on Signals and Images via the Fourier Transform, Genomics, EHRs, GPUs, and Next Generation Computational Statistics

OpenAlex ID: W2891764468

Cites: 28

Cited by:

Related to: 10