Depending on whom you ask reml stands for residual

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: for these special cases and produces less biased estimates than ML in general. Depending on whom you ask, REML stands for REsidual Maximun Likelihood or REstricted Maximum Likelihood. Copyright c 2012 (Iowa State University) Statistics 511 7 / 31 The REML Method 1 Find n − rank(X) = n − r linearly independent vectors a1 , . . . , an−r such that ai X = 0 for all i = 1, . . . , n − r. Find the maximum likelihood estimate of γ using w1 ≡ a1 y, . . . , wn−r ≡ an−r y as data. w1 a1 y .. A = [a1 , . . . , an−r ] w= . = . =Ay . . wn−r an−r y 2 Copyright c 2012 (Iowa State University) Statistics 511 8 / 31 If a X = 0 , a y is known as an error contrast. Thus, w1 , . . . , wn−r comprise a set of n − r error contrasts. Because (I − PX )X = X − PX X = X − X = 0, the elements of (I − PX )y = y − PX y = y − ˆ y are each error contrasts. Copyright c 2012 (Iowa State University) Statistics 511 9 / 31 Because rank(I − PX ) = n − r, there exists a set of n − r linearly independent rows of I − PX that can be used in step 1 of the REML method to get a1 , . . . , an−r . If we do use a subset of rows of I − PX to get a1 , . . . , an−r ; the error contrasts w1 = a1 y, . . . , wn−r = an−r y will be a subset of the elements of the residual vector (I − PX )y = y − ˆ. y This is why it makes sense to call the procedure Residual Maximum Likelihood. Copyright c 2012 (Iowa State University) Statistic...
View Full Document

This document was uploaded on 03/27/2014.

Ask a homework question - tutors are online