Least square regression

[mattc]

hat do we mean by the "least square" in regression analysis procedure? Where do you use this procedure?

[omega]

We minimize the sum of the squares of the residuals of the data points from the curve, ie. we minimize sum[(yi-yihat)^2], where yi is the ith observation, and yihat is the ith fitted value from the line.

We use this procedure in regression and we can fit regression line, polynomials, and multiple regression.