Levenberg-Marquardt algorithm is a gradient-based, deterministic local optimization algorithm. The Levenberg-Marquardt algorithm has an advantage over the traditional Back Propagation algorithm, where it can provide faster (second-order) convergence rate and keep relative stability