1 pr oc sprocesso w y n x dados es in 32 gerador

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: y (3.1) pr oc• sProcesso ρW y + α ι n + X dados: θ + ε es in (3.2), = Gerador de β + W X y = ( I n − ρW ) − 1 ( α ι n + X β + W X θ + ε) ( 3.2) 2 ε ∼ N (0, σy I+ ) ι + X β + W X θ + ε y = ρW n α n (3.1) y n ( 1 v ρ t o) − 1 α o + n β n represent ) where 0 represent s an = ×I n − ecW r of (zeιrn s aXd ι+ W X θ + εs an n × 1 vec( 3.2) t or of ones associ at ed wε t∼ N (0, onI nant t erm paramet er α . T hi s model can be i h t he cσ2 st ) wr it t en as a SAR model by defining: Z = ι n X W X and δ = α β θ , where 0 3.3) . T t is m n × t vec h of zer o ood ι n represen s an a× veM leading t o (represenhs an eans1 hat ttore likelihs andfun ct ion for t SA R n nd1SDct or moofelones n beci at iedewi ti h t he const anmodelos SARαe SDM otdel can b e d s• ca asso wr t t n n t he same f or m whparamet = ι.n T hi sfor éhe SAR A função LL, para os t t erm er e: Z er X m a W X fo mowr it aenmesmo,X model byr defining: Z = el. n X W X and δ = α β θ , del t nd as = SARveremos:t he SDM mod ι Z a ιn leading t o ( 3.3) . T his means t hat t he likelihood f unct ion...
View Full Document

{[ snackBarMessage ]}

Ask a homework question - tutors are online