On a Consistent Rank Estimate in a Linear Structural Model
2012 (English)In: PROBASTAT '11: Proceedings of the Sixth International Conference on Probability and Statistics: Dedicated to Professor Lubomir Kubacek in Recognition of His Eightieth Birthday, 2012, 191-202 p.Conference paper (Refereed)
The structural linear model is considered that is an errors-in-variables model where the unobserved variables are i.i.d. In this model we can find linear transformations depending on the parameter, such that the transformed observations using the true parameter are uncorrelated. Then a parameter estimator is defined as a zero point of a consistent correlation estimator. A rank estimation is proposed as a zero point of Kendall's correlation measure and its consistency is shown. While the Pearson estimate of the covariance delivers the total least squares estimate.
Place, publisher, year, edition, pages
2012. 191-202 p.
, Tatra Mountains Mathematical Publications, ISSN 1210-3195 ; 51
rank estimate, structural linear model, Kendall's tau, total least squares, Theil-Sen, estimate
IdentifiersURN: urn:nbn:se:uu:diva-196152DOI: 10.2478/v10127-012-0019-7ISI: 000314254900019OAI: oai:DiVA.org:uu-196152DiVA: diva2:609283
6th International Conference on Probability and Statistics (PROBASTAT 2011), JUL 04-08, 2011, Smolenice, SLOVAKIA