For the mathematical-programming problem: Minimize f (x), subject to {g_i(xv) ≤ b_i}, the saddle-point problem is to find vectors xv⁰ and yv⁰ such that F (xv ⁰, yv) ≤ F (xv⁰, yv ⁰) ≤ F (x, yv ⁰), where F (x, yv) is the associated Lagrangian function, yv ≥ 0.Saddle-point of a function.