The use of methods of entropy and decline of The character in the estimation of linear regression parameters Specimen existence of linear multi- function problem | ||
Al Kut Journal of Economics Administrative Sciences | ||
Article 1, Volume 0, Issue 22, August 2018, Pages 221-236 | ||
Abstract | ||
Abstract Many research and studies concerned with the problem of diversity in terms of linear inception and methods disclosed in addition to the effects in the process of assessing aspecimen linear regression parameters when random errors follow a normal distribution, as well as For the development of appropriate solutions to reduce the effects of that problem. Using some of the estimation methods that rely on entropy (function Entropy) It is a method of public Great Entropy (GME) (Generalized Maximum Entropy Method) and is destined to Leuven (first) (MELE1) (Maximum Entropy Leven 1 Estimator), has been compared way downhill normal character (ORR) (Ordinary Ridge Regression Method) where the use of standard Average boxes error (MSE) to compare these methods. ــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــــ (*) بحث مستل من رسالة ماجستير للباحث الثاني. The comparison was made between these methods to data generated using simulation method through system programming (R), where data has been generating a form to decline by four explanatory variables and different sizes samples (n = 20,60,150) And various links (0.85,0.90,0.99), and the simulation results indicate that the best methods and capabilities are the way public Great entropy (GME) and then the usual way downhill character (ORR), where the differences between them are very slim. | ||
Statistics Article View: 205 PDF Download: 182 |