Change search

Algorithms for robustified error-in-variables problems
Luleå University of Technology, Department of Engineering Sciences and Mathematics, Mathematical Science.
1998 (English)In: COMPSTAT [1998]: proceedings in computational statistics ; 13th symposium held in Bristol, Great Britain, 1998 ; with 61 tables / [ed] Roger Payne; Peter Green, Heidelberg: Physica-Verlag Rudolf Liebig GmbH , 1998, 293-298 p.Conference paper (Refereed)
##### Abstract [en]

From the introduction: We consider the problem of fitting a model of the form $y=f(x,\beta)$ to a set of points $(x_i,y_i)$, $i=1,\dots,n$. If there are measurement or observation errors in $x$ as well as in $y$, we have the so-called errors-in-variables-problem with model equation $$y_i=f(x_i+\delta_i,\beta)+\varepsilon_i,\ i=1,\dots,n,\tag 1$$ where $\delta_i\in\bbfR^m$, $i=1,\dots,n$, are the errors in $x_i\in\bbfR^m$. Then the problem is to find a vector of parameters $\beta\in\bbfR^p$ that minimizes the errors $\varepsilon_i$ and $\delta_i$ in some loss function subject to (1). We present algorithms using more robust alternatives to the least squares criterion.\par We will further discuss, from

##### Place, publisher, year, edition, pages
Heidelberg: Physica-Verlag Rudolf Liebig GmbH , 1998. 293-298 p.
##### Research subject
Scientific Computing
##### Identifiers
Local ID: cbd8b9b0-88b1-11dd-9d47-000ea68e967bISBN: 3-7908-1131-9OAI: oai:DiVA.org:ltu-38368DiVA: diva2:1011868
##### Conference
COMPSTAT : 24/08/1998 - 28/08/1998
##### Note
Godkänd; 1998; 20080922 (ysko)Available from: 2016-10-03 Created: 2016-10-03Bibliographically approved

No full text

#### Search in DiVA

##### By author/editor
Edlund, OveEkblom, Håkan
##### By organisation
Mathematical Science