Use Matlab to simulate an ARMA(p,q) process and then use the Durbin-Levinson and Innovation Algorithms to do one-step ahead prediction. Run your code for different values of sigma^2, the variance of the driving white noise process.


In [1]:
arma.sim<-arima.sim(model=list(ar=c(.9,-.2),ma=c(-.7,.1)),n=100)

In [2]:
ts.plot(arma.sim)



In [3]:
arma.acf<-acf(arma.sim,type="correlation",plot=T)



In [ ]: