#autocorrelation

Question 1: Select all that apply

You are choosing lag features for the monthly US unemployment rate, whose autocorrelation is 0.99 at lag 1 and 0.98 at lag 2. Which statements are true? Select all that apply.

The lag-2 autocorrelation may only be lag-1 information carried over, so by itself it does not justify a lag-2 feature1
The partial autocorrelation of a lag is its correlation with the series after accounting for all shorter lags2
Autocorrelation and partial autocorrelation measure only linear dependence, so a lag plot can show relationships they miss3
Every lag whose autocorrelation is above 0.5 should become a feature4
A lag plot draws the series and its lagged copy as two lines against time5

Timed test