Accalerated Back-propagation Learning: Two Optimization Methods