In linear modeling of continuous variables, the basic equation is

=b₀ +b₁x₁ + b₂x₂ + ...+ b_px_p

The goal is to find beta-weightsb₀, ... , b_p so that

a) The differences between predicted values and true values-y are as small as possible.

b) Theand y match exactly

c) The sum of squared deviations (-y )²is minimized.