How to get the MA coefficients for ARMA model given a timeseries dataset

timeseries
arma

#1

MY dataset is as below:

image

Can anyone please explain how to get AR and MA coefficients for this dataset??
Ur help is appreciated.Thnx in advance.


#2

I’m assuming that you’re a R user and already know about the forecast package.

So the simplest way is to perform the auto.arima on your TS (time-series object made using ts() from forecast package) object and then run $coef . This will give you the coefficients based on system selected ARIMA order (p,d,q).

You can perform the same steps with ARMA model as well and get your coefficients.

Let me know if this solves your problem.

Peace.


#3

No…i am not using R language nor the package. I want to know how the coefficients are computed given a dataset step-by-step. With an example it would be great…


#4

If you want to look the ARMA process in isolation, it’s nothing but linear regression of the dependent variable with it’s own lagged values till order “p”. The process of coefficient calculation is similar to what is in Linear regression. For more details, you can refer AR, MA and ARMA models


#5

The best scientific approach is by plotting the ACF and PACF and actually counting the number of lines (lags) above the threshold. This approach helps as you can see the changes in ACF and PACF while you incrementally try to get there.

Auto.Arima though it provides a good estimate i usually prefer bottoms up starting at 0,0 0 and slowly adding and checking the ACF and PACF plots of the residual.


#6

thanx @gauravpandey.pgpm17c for the ideal single word answer!..Would you mind explaining the arma in comparison to linear regression. An example on a dataset will help. The link is lengthy and not easy to pick up. It incorporates multiple case scenarios which I am not interested. I am looking for a simple normal implementation of the ARMA.

Thnx again


#7

Can you explain how is coefficient estimation for MA(2) model and obtain \theta_1 and \theta_2. . I heard using MLE its done. could you explain the same in layman terms? I am stuck with this for soo long…No sources explain in simple terms. Please help