W v es n inuenceyon explodeptet o c tua ar iabtle i t

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: sitaptture ulsing Autoregressivescenar io involving )nt partir influences. W v es n influenceyon explodeptet o c tua ar iabtle. I t ais eressof l at e t a are r eadi l eavai uma nden v ion espacial unlikimpleha o agente que toma t he bl dependência hese t yp y ely t entre temporal e a ehisexistência busing a i er o simple ecenanio in t lving pe dent v ioa de y h il pl e y sx ai r ed b olr edtarenrsit uatar avaieabtl eattovsccymuret elheseptlyp es ofyvlowontxpl anator y var i esl es x aanecisãonh seu oci at ed aptl ar par amet er s β and at. eFor si mpl i ci t y, we eadizywi te assvizinho. ca l l t dd ab s θ e y rte t hisssitomtpl etel y ex pl ai ned simpleoscenaaioainvolvinghat i c ua ion using a ver y by t w ex pl r n t or y var i plo assume t he n × 1 vect or s x and z are di st ri but ed N (0, I n ), and we assume x alssoctzhtatdi s calmrnldenVariávelsOmitida.θ. oForplsi nat or yt y, rwe e d e is Viés de r y ex ea β b d any i aare 2. dop e pt eltametplr i nedany t w ex ampl i ci va i n e ap ea . ctasr s ci aand scal ar dartamet terd β andI n .), Fordsiwe lasstumwe oso x t e d z are pi s ri bu es N (0, θ an mp i ci y, e x nt . vect or s x andVariáveisst ri but ed y = modelos Espaciais: me x N (0, I ), θ (2.7) • z are di omitidas em x βn+ zand we assu amenidades locacionais, dent . acesso a rodovias, bairros tradicionais, etc. Given bot h x and z, sol ut ion of t he linear syst em would yield an exact β and y = β + ur θ. T he absence ofxa di stzθ bance t er m si mpl i fi es di scover y(2.7) e par amet er s. of t h y caseβ + zθt he vect or z is not obser ved. Since t he unobser ved = x where (2.7) Consider t he solrut bl e zof t het lineaelatyst em h tould byield anect orct ,βwancan st i l l uncover ion is no cor r r s ed wit w he o ser ved v exa x e d va i a z,i stol ut ioncof tter m si mr lsysts di scould y of t he pxaametand. s ur a e linea p eem yield an ea ct β e dβ . I n bhin casehehe vect oirfizθ actw ovstãoeindependentesee a,...
View Full Document

Ask a homework question - tutors are online