These are algorithms for convex optimization whereas stochastic gradient descent and saddle free newton are generally used for nonconvex optimization (neural ne

These are algorithms for convex optimization whereas stochastic gradient descent and saddle free newton are generally used for nonconvex optimization (neural nets in particular).