Part 4

Question 1

White noise

Answer 1

White noise is a sequence of random variables with a mean of zero, finite/constant variance and no autocorrelation at any lag. Each variable is uncorrelated with others making the sequence unpredictable

Question 2

Model selection would typically not be done using ACFs. True or False

Answer 2

True

Question 3

Which information criterion should be preferred if they suggest different model orders? (AIC, BIC and HQ)

Answer 3

• AIC is not consistent, will typically pick “bigger” models.
• BIC is strongly consistent but inefficient.
• HQ is in between