You are on page 1of 3

8/5/2018 Practical Time Series Analysis - Home | Coursera

 ARMA Processes 3/3 points (100.00%)


Quiz, 3 questions

 Congratulations! You passed! Next Item

1/1
 points

1.
Recall our framework:

Xt = N oise + Autoregressive P art + M oving Average P art

Xt = Zt + ϕ1 Xt−1 + … + ϕp Xt−p + θ1 Zt−1 + … + θq Zt−1

Θ(B)Zt = Φ(B)Xt

Θ(B)
Ψ(B) = Φ(B)

Θ(B)
Xt = Z
Φ(B) t
= Ψ(B)Zt

We develop an MA model for the mixed process

Xt = 0.5Xt−1 + Zt + 0.2Zt−1 .

Find Ψ(B) and use it to express the denominator as a geometric series.

1 + .5B + .25B 2 + .125B 3 + ⋯

Correct
Yes, that's terri c! We use

1+.2B
Ψ(B) = 1−.5B

and then express

1
1−.5B
= 1 + (.5B) + (.5B)2 + (.5B)3 + …

1 − (.5B) + (.5B)2 − (.5B)3 + …

1/1
 points

2.

https://www.coursera.org/learn/practical-time-series-analysis/exam/Sj5fq/arma-processes 1/3
8/5/2018 Practical Time Series Analysis - Home | Coursera

Now multiply

 ARMA Processes
Ψ(B) = (1 + .2B)(1 + .5B + .25B 2 + .125B 3 + ⋯) 3/3 points (100.00%)
Quiz, 3 questions
and obtain an expression for Xt :

Ψ(B) = 1 + .7B + .35B 2 + .175B 3 + …

Correct
Good work. We multiply just as we regularly do for nite order polynomials.

Ψ(B) = 1 + .5B + .2B 2 + …

1/1
 points

3.
Finally, which ACF looks to be consistent with this process?

ACF1:

https://www.coursera.org/learn/practical-time-series-analysis/exam/Sj5fq/arma-processes 2/3
8/5/2018 Practical Time Series Analysis - Home | Coursera

 ARMA Processes
ACF2: 3/3 points (100.00%)
Quiz, 3 questions

Correct
Good! The coe cients on the polynomial should decay rather quickly.

  

https://www.coursera.org/learn/practical-time-series-analysis/exam/Sj5fq/arma-processes 3/3

You might also like