Enviar registro por correo electrónico: Reinforcement learning and dynamic programming using function approximators