Advances in large scale unconstrained optimization: novel preconditioning strategies for Nonlinear Conjugate Gradient methods and new developments in Newton-Krylov methods