Training Neural Networks Using the Property of Negative Feedback to Inverse a Function