Talk:Moving-average model


estimation: how?

edit

For AR(p), we can estimate it by solve the Yule-Walker equations, but for an MA(q) process, how do we estimate its parameters? AR(1) in error: estimated generalized least square or lagged variables or distributed lags as shown in http://www.xycoon.com/autocorrelation.htm

Jackzhp (talk) 02:37, 11 April 2009 (UTC)Reply

Clarification needed: what is meant by 'additional interpretation'?

edit

In the 'Interpretation' section, what is the "additional interpretation"? "The moving-average model is essentially a finite impulse response filter applied to white noise, with some additional interpretation placed on it." Afonso Dimas Martins (talk) 20:06, 27 February 2023 (UTC)Reply