Abstract
We will study a family of methods with the same theoretical and practical properties for quadratics as the conjugate gradient method. This family is a convex combination of the objective function and the squared norm of the function gradient. We will also analyze the possible practical advantages when extended to strictly convex functions.(AU)