00:01
All right, here we have why are errors squared in a regression? all right, so the errors in a regression is the difference between the observed value and the values predicted by the model.
00:12
Okay, so it means that if we take our observed model, observed, right, minus our predicted, right, predicted, predicted.
00:24
That's going to give us positive and negative values depending on whether or not what is observed is higher or lower than the prediction.
00:36
So it's going to give us positive or negative values.
00:38
Now here's the key.
00:39
If we combine them just as is like that, then those positive and negative values are going to cancel out.
00:46
Okay, so the reason why we do it is because summing, as i just said, summing positive and negative errors will cancel them out...