A procedure for minimizing an objective function that is a sum of squares. The method is similar to the Newton Raphson method but with the advantage of not requiring second derivatives of the function.