Skip to main navigation Skip to search Skip to main content

Statistical Curvature in Two Dimensions: A Teaching Application

  • Bernard Ellem
  • , Kylie Thomas

    Research output: Book/ReportReportOther

    Abstract

    This report presents a novel method of teaching curvature measures via simple 2D examples using the elementary formulae for curvature encountered by undergraduate students. Using this procedure students can verify algebraically and numerically the invariance of intrinsic curvature and corroborate the best parameterisation by examination of parameter effects curvature, again mathematically and empirically. For serious users of curvature measures, this elementary exposition also reconciles the general definition of 'statistical curvature' coined by Efron (1975), with the approach of Amari (1990), albeit for the exponential connection only.
    Original languageEnglish
    Place of PublicationArmidale, Australia
    PublisherUniversity of New England, School of Mathematics, Statistics and Computer Science
    Publication statusPublished - 2005

    Publication series

    NameResearch and Technical Reports

    Keywords

    • Statistics

    Fingerprint

    Dive into the research topics of 'Statistical Curvature in Two Dimensions: A Teaching Application'. Together they form a unique fingerprint.

    Cite this