Home > Research > Publications & Outputs > Linex and double-linex regression for parameter...

Electronic data

  • linex_REVISED_3

    Accepted author manuscript, 709 KB, PDF document

    Embargo ends: 22/12/23

    Available under license: Other


Text available via DOI:

View graph of relations

Linex and double-linex regression for parameter estimation and forecasting

Research output: Contribution to Journal/MagazineJournal articlepeer-review

<mark>Journal publication date</mark>30/04/2023
<mark>Journal</mark>Annals of Operations Research
Issue number1-2
Number of pages17
Pages (from-to)229-245
Publication StatusPublished
Early online date22/12/22
<mark>Original language</mark>English


The choice of an estimation method has received considerable attention in the Operations Research literature. In this paper we depart from the standard use of linex and double-linex loss functions which are widely used in parameter estimation and forecasting problems and we propose a non-standard use for them. Specifically, we propose to use the corresponding linex and double-linex error densities as models for the errors of a regression problem when more emphasis should be placed on over-estimation or under-estimation of errors. The new techniques are applied to synthetic as well real data concerning the role of management in production as well as to an application of forecasting volatility in intradaily data.