Regularized Difference Criterion for Computing Discriminants for Dimensionality Reduction

Alex J. Aved, Erik P. Blasch, Jing Peng

Research output: Contribution to journalArticleResearchpeer-review

2 Citations (Scopus)

Abstract

Hyperspectral data classification has shown potential in many applications. However, a large number of spectral bands cause overfitting. Methods for reducing spectral bands, e.g., linear discriminant analysis, require matrix inversion. We propose a semidefinite programming for linear discriminants regularized difference (SLRD) criterion approach that does not require matrix inversion. The paper establishes a classification error bound and provides experimental results with ten methods over six hyperspectral datasets demonstrating the efficacy of the proposed SLRD technique.

Original languageEnglish
Article number7906579
Pages (from-to)2372-2384
Number of pages13
JournalIEEE Transactions on Aerospace and Electronic Systems
Volume53
Issue number5
DOIs
StatePublished - 1 Oct 2017

Fingerprint

Discriminant analysis

Keywords

  • Classification
  • RELIEF
  • dimensionality reduction
  • feature selection
  • hyperspectral imaging

Cite this

@article{3a71db48a29443a7a220194471fac952,
title = "Regularized Difference Criterion for Computing Discriminants for Dimensionality Reduction",
abstract = "Hyperspectral data classification has shown potential in many applications. However, a large number of spectral bands cause overfitting. Methods for reducing spectral bands, e.g., linear discriminant analysis, require matrix inversion. We propose a semidefinite programming for linear discriminants regularized difference (SLRD) criterion approach that does not require matrix inversion. The paper establishes a classification error bound and provides experimental results with ten methods over six hyperspectral datasets demonstrating the efficacy of the proposed SLRD technique.",
keywords = "Classification, RELIEF, dimensionality reduction, feature selection, hyperspectral imaging",
author = "Aved, {Alex J.} and Blasch, {Erik P.} and Jing Peng",
year = "2017",
month = "10",
day = "1",
doi = "10.1109/TAES.2017.2696236",
language = "English",
volume = "53",
pages = "2372--2384",
journal = "IEEE Transactions on Aerospace and Electronic Systems",
issn = "0018-9251",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "5",

}

Regularized Difference Criterion for Computing Discriminants for Dimensionality Reduction. / Aved, Alex J.; Blasch, Erik P.; Peng, Jing.

In: IEEE Transactions on Aerospace and Electronic Systems, Vol. 53, No. 5, 7906579, 01.10.2017, p. 2372-2384.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Regularized Difference Criterion for Computing Discriminants for Dimensionality Reduction

AU - Aved, Alex J.

AU - Blasch, Erik P.

AU - Peng, Jing

PY - 2017/10/1

Y1 - 2017/10/1

N2 - Hyperspectral data classification has shown potential in many applications. However, a large number of spectral bands cause overfitting. Methods for reducing spectral bands, e.g., linear discriminant analysis, require matrix inversion. We propose a semidefinite programming for linear discriminants regularized difference (SLRD) criterion approach that does not require matrix inversion. The paper establishes a classification error bound and provides experimental results with ten methods over six hyperspectral datasets demonstrating the efficacy of the proposed SLRD technique.

AB - Hyperspectral data classification has shown potential in many applications. However, a large number of spectral bands cause overfitting. Methods for reducing spectral bands, e.g., linear discriminant analysis, require matrix inversion. We propose a semidefinite programming for linear discriminants regularized difference (SLRD) criterion approach that does not require matrix inversion. The paper establishes a classification error bound and provides experimental results with ten methods over six hyperspectral datasets demonstrating the efficacy of the proposed SLRD technique.

KW - Classification

KW - RELIEF

KW - dimensionality reduction

KW - feature selection

KW - hyperspectral imaging

UR - http://www.scopus.com/inward/record.url?scp=85023195874&partnerID=8YFLogxK

U2 - 10.1109/TAES.2017.2696236

DO - 10.1109/TAES.2017.2696236

M3 - Article

VL - 53

SP - 2372

EP - 2384

JO - IEEE Transactions on Aerospace and Electronic Systems

JF - IEEE Transactions on Aerospace and Electronic Systems

SN - 0018-9251

IS - 5

M1 - 7906579

ER -