Principal Component Analysis and Geodetic Networks

des Instituts für Geodäsie

Heft 7/1982

Principal Component Analysis
and Geodetic Networks
-  Some Basic Considerations  -

Calgary, Alberta, Canada

On leave from the University of Hannover,
Federal Republic of Germany


In: BORRE, Kai / WELSCH, Walter M. (Eds.) [1982]:
International Federation of Surveyors - FIG -
Proceedings Survey Control Networks
Meeting of Study Group 5B, 7th-9th July, 1982, Aalborg University Centre, Denmark

Schriftenreihe Wissenschaftlicher Studiengang Vermessungswesen, Hochschule der Bundeswehr München, Heft 7, Neubiberg, S. 275-291.


As additional criteria for the diagnosis of geodetic networks, principal component analysis (PCA) is introduced and its geometrical and statistical properties are outlined. These criteria show geometrically the correlation between variates, i.e. information that cannot always be derived from other common criteria for precision. As the examples will show, PCA yields a geometrically clear insight into the network behaviour and it is therefore ideally suited to the description of certain global properties of a network, especially one designed for monitoring.

zurück zur Übersicht des Heftes 7






(836 KB)