This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: var(delta + wt) = var(wt) E(delta+wt) = delta B*Xt = Xt-1 B^2 * Xt = B(B Xt) = Xt-2 delta = 1 - B delta(delta(xt)) = (1-B)[(1-B)(xt)) = (1-2B+B^2)*Xt = Xt - 2Xt-1 + Xt-2 In R, use diff for differencing delta(constant B0) = 0------------gamma(h) = cov(xt, t+h) gamma(0) = variance(xt) p(h) = autocorrelation btwn xt and xt+h p(h) = cov(xt, xt+h) / stdev(xt)*stdev(xt+h) but the 2 stdev are the same so it’s the variance on the bottom ph() = gamma(h) / gamma(0)[email protected] differencing might introduce additional dependencies that makes the model useless AR and MA models capture some type of dependency---------------...
View Full Document
- Fall '13
- Autoregressive moving average model, Xt, Autoregressive model