Figure 2.18:
Figure 2.18
The least-squares idea: make the errors in predicting
y
as small as possible by minimizing the sum of their squares.
