Key points are not available for this paper at this time.
Multiple regression, one form of estimation by least squares, is an important statistical technique. Heavy computing is involved and a great many programs have been written to do the necessary work—so many that it is clear that most of them do not attain an adequate degree of generality. This paper outlines the essential features of regression analysis and attempts to give the essential requirements for a general program.
M. J. R. Healy (Tue,) studied this question.