English

Error structures and parameter estimation

Statistics Theory 2007-06-13 v1 Functional Analysis Probability Statistics Theory

Abstract

This article proposes a link between statistics and the theory of Dirichlet forms used to compute errors. The error calculus based on Dirichlet forms is an extension of classical Gauss' approach to error propagation. The aim of this paper is to derive error structures from measurements. The links with Fisher's information lay the foundations of a strong connection with experiment. We show that this connection behaves well towards changes of variables and is related to the theory of asymptotic statistics.

Keywords

Cite

@article{arxiv.math/0612258,
  title  = {Error structures and parameter estimation},
  author = {Nicolas Bouleau and Christophe Chorro},
  journal= {arXiv preprint arXiv:math/0612258},
  year   = {2007}
}