Question

2. Correlograms are useful for modelling time series: The 4 plots on the next page show the correlograms of 4 time series. These correlograms are estimated based on 196 observations. Which of these series is likely to be white noise? For each of the series that are not white noise, determine the order of the AR model which would be suitable for that series

Answer 1

there are many types of AR model like............

1)AR (P) model.......

perameter would be -1 ><1

2)Moving average model

in that ats are the white noise....and $\Theta$1,$\Theta$2...are the perameters.....q is called order for the model.

so Zt=at-$\Theta$1 *at-1...............$\Theta$q*at-q

3) ARMA (AUTOREGRESSIVE MOVING AVERAGE MODEL)

here we can have both noise and non noise perameters.