ParrotPete
Oct12-11, 02:59 AM
Hello. I am given an ARMA process that looks like this.
Y[n = b0Xn+b1X(n-1)
Where Xn is a gaussian white noise process. N(0,\sigma2 Now there is the general way to solve linear difference equations. But would it be possible to just use Z-transform to obtain a closed formula for Yn, and then argue that Y is the sum of gaussian variables, and therefore gaussian?
Y[n = b0Xn+b1X(n-1)
Where Xn is a gaussian white noise process. N(0,\sigma2 Now there is the general way to solve linear difference equations. But would it be possible to just use Z-transform to obtain a closed formula for Yn, and then argue that Y is the sum of gaussian variables, and therefore gaussian?