An ascent method for unconstrained optimization such that successive directions are conjugate with respect to the hessian for a quadratic objective function.

Category: Nonlinear Programming