Approximate Parametrisation of Confidence Sets

  • Zbyněk Šír
Conference paper


In various geometrical applications, the analysis and the visualization of the error of calculated or constructed results is required. This error has very often character of a nontrivial multidimensional probability distribution. Such distributions can be represented in a geometrically interesting way by a system of so called confidence sets. In our paper we present a method for an approximate parametrisation of these sets. In sect. 1 we describe our motivation, which consists in the study of the errors of so called Passive Observation Systems (POS). In sect. 2 we give a result about the intersection of quadric surfaces of revolution, which is useful in the investigation of the POS. In sect. 3 we give a general method for an approximate parametrisation of the confidence sets via simultaneous Taylor expansion. This method, which can be applied in a wide range of geometrical situations, is demonstrated on a concrete example of the POS.


Taylor Expansion Observation Site Multivariate Normal Distribution Multivariate Distribution Quadric Surface 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Wiley, R.G.: Electronic Intelligence: The Interception of Radar Signals, Artech House 1985, 284 p.Google Scholar
  2. 2.
    Mizera, I.: On depth and deep points: A calculus, Ann. Stat. 30(2002), pp 1681–1736zbMATHMathSciNetCrossRefGoogle Scholar
  3. 3.
    Johnson, N.L.; Kotz, S.; Balakrishnan, N.: Continuous multivariate distributions, Vol. 1: Models and Applications, New York, Wiley 2000, 752 p.Google Scholar
  4. 4.
    Walter, W.: Analysis 2, Grundwissen Mathematik Bd. 4, Springer-VErlag 1990, 396p.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2005

Authors and Affiliations

  • Zbyněk Šír
    • 1
  1. 1.Charles UniversityPragueCzech Republic

Personalised recommendations