Holt's Exponential Smoothing
Back Practice Assignment • 10 min
English
Graded Assignment
1.
Holt's Exponential Smoothing
If our Holt’s Exponential Smoothing model is performing well for a trending data set, what
will the plot of the within-sample forecast look like when plotted on the same axes as the
1 point
observed data?
Review Learning Objectives
The plot of the forecasts should be mostly parallel to the observed data and be
consistently over-estimating the observed data.
The plot of the forecasts should be roughly a horizontal line that goes through the
Ready to review
middle what you’ve
of the observed datalearned
set. before starting the assignment? I’m here to help.
The Help
plot of
methe forecasts should
practice closely
Let's chat follow the observed data, following its trend
by neither consistently over-estimating nor consistently under-estimating the
observed data.
The plot ofdetails
Assignment the forecasts should be mostly parallel to the observed data and be
consistently under-estimating the observed data.
Attempts
Unlimited Resume
2. For Holt’s Exponential Smoothing, what is beta? 1 point
Beta is a weight for the Trend equation, and beta is between minus one and zero.
Your grade
Beta
You is a weight
haven’t forthis
submitted theyet.
Trend equation,
We keep and beta
your highest [Link] between zero and one.
Beta is a weight for the Trend equation, and beta is between zero and 100.
--
Beta is a weight for the Trend equation, and beta is between minus one and plus
one.
Like Dislike Report an issue
3. Intuitively what does the Trend equation for Holt’s Exponential Smoothing essentially 1 point
convey?
The trend value for this period is equal to beta times the previous level plus one
minus beta times the difference in the trend (of this period and the last period).
The trend value for this period is equal to beta times the difference in the level (of
this period and the last period) plus one minus beta times the difference in the trend
(of this period and the last period).
The trend value for this period is equal to beta times the previous level plus one
minus beta times the previous trend.
