This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: EXERCISE: The Identification of ARMA Models An appropriate ARMA model for fitting to a stationary series may be identified by inspecting the following three functions: (1) The Empirical Autocorrelation Function, (2) The Empirical Partial Autocorrelation Function, (3) The Nonparametric Estimate of the Spectral Density Function. The procedures for determining the orders of an ARMA process from the first two of these functions described in the appendix to Lecture 4 which is titled Identification of ARIMA Models . Some practice in identifying the orders can be gained by using the TSERIES program and the MESOSAUR program. The Pseudo-Random Data Series Within the TSERIES program, there are facilities for generating pseudo-random data from specified models. You are invited to generate data from each of the models specified in the following list under the numerals (i) to (xiv). In the process, you may consider the effects of varying the sample size. Alternatively, data that have been generated in this manner by...
View Full Document
- Spring '12
- spectral density, Autocorrelation, Stationary process, Autoregressive moving average model, Time series analysis, ARMA Models