Google+

Least-square procedure

It is technique for fitting a straight line to a scattergram in regression. This technique determines the best-fitting line by minimizing the square of the vertical distance of all the points from the line. The best-fitting line is called the regression line.