In regression models, outliers, or results with large errors, are almost always present. As a result, it's critical to investigate the regression with as few outliers as feasible. So, in order to address this issue, I investigated a novel resilient Least Square (LS) Method based on a genuine stochastical model and dubbed after the author PEROBLS. In other words, the variance of an observation with a gross mistake is replaced by the mean square error, which is made up of both random and gross observation error. The gross error is precisely and unbiasedly predicted in the LS model, therefore only one or two correction iterations are required.
Author(S) Details
G. Perovic
Department of Geodesy and Geoinformatics, Faculty of Civil Engineering, University of Belgrade, Serbia.
View Book:- https://stm.bookpi.org/RDST-V5/article/view/6871
No comments:
Post a Comment