Abstract
Canonical discriminant analysis (CDA) and linear discriminant analysis (LDA) are popular classification techniques. Likewise, practitioners, who are familiar with regularized discriminant analysis (RDA), soft modeling by class analogy (SIMCA), principal component analysis (PCA), and partial least squares (PLS) will often use them to perform classification. In this chapter, we will attempt to make some sense out of all of this. We will explain when CDA and LDA are the same and when they are not the same. We will also discuss the relative merits of the various stabilization and dimension reducing methods used, focusing on RDA for numerical stabilization of the inverse of the covariance matrix and PCA and PLS as part of a two-step process for classification when dimensionality reduction is an issue.
Original language | English |
---|---|
Title of host publication | Comprehensive Chemometrics |
Subtitle of host publication | Chemical and Biochemical Data Analysis, Second Edition: Four Volume Set |
Pages | 585-603 |
Number of pages | 19 |
Volume | 3 |
ISBN (Electronic) | 9780444641656 |
DOIs | |
State | Published - Jan 1 2020 |
Bibliographical note
Publisher Copyright:© 2020 Elsevier B.V. All rights reserved
Keywords
- SIMCA
- canonical discriminant analysis
- discriminant PLS
- linear discriminant analysis
- oriented PLS
- quadratic discriminant analysis
- regularized discriminant analysis
ASJC Scopus subject areas
- General Chemistry