00:01
So what we have right here are the actual values and the forecasted values of the time series.
00:08
And we have a researcher who wants to evaluate a fit of the estimated equation to the time series.
00:14
And so what this researcher wants to know is the mad, which is the mean absolute difference, and the sum of squares error, sse or see, sum of squares of error.
00:32
You might even see it as sse, sum of squares of the error.
00:36
Either way, we're talking about the sum of the squares.
00:39
So let's go ahead and figure out the mad first.
00:40
So the mean absolute difference.
00:42
So absolute means your absolute value.
00:45
So you take your actual value minus your forecasted value, and then you divide it by how many there are.
00:59
So it's the mean absolute difference.
01:02
So there's that.
01:03
And then to get the sum of squares error, what we do is we take the sum of each of the actual value minus your forecast, and you square that.
01:32
You square that difference and you sum all those up.
01:39
So the mean absolute difference here is the absolute differences.
01:43
195 minus 87 is eight, 162 minus seven is 13, and so on and so forth...