Principal Component Analysis: Basic Methods and Extensions,

Abstract

The basic method of principal component analysis is relatively well understood by physical oceanographers. Some less generally understood ideas involve significance testing and rotation of the basis functions. Also, a number of other analysis techniques related to principal component analysis can be more easily understood by using it as a starting point. Such techniques include factor analysis, extended empirical orthogonal function analysis, canonical correlation analysis, and complex empirical orthogonal function analysis, for example. The basis calculations comprising principal component analysis are presented, and significance testing and rotation are discussed. Pacific sea level data are used to illustrate these techniques. The paper concludes with a discussion of various extensions to the basic technique and an evaluation of the usefulness of the extensions.

Document Details

Document Type
Technical Report
Publication Date
Nov 01, 1993
Accession Number
ADP008734

Entities

People

  • Gary T. Mitchum

Organizations

  • University of HawaiĘ»i System

Tags

DTIC Thesaurus Topics

  • Computing-Related Activities
  • Correlation Analysis
  • Data Science
  • Factor Analysis
  • Information Science
  • Interdisciplinary Science
  • Mathematical Analysis
  • Mathematics
  • Oceanography
  • Physical Oceanography
  • Rotation
  • Sea Level
  • Statistical Analysis
  • Statistics

Readers

  • Approximation Theory.
  • Theoretical Analysis.