What is neural net training?
Networks first need to be trained by being presented with hundreds or even thousands of facts, each fact consisting of inputs and corresponding outputs. Through a unique feedback process, the network learns how those inputs are related to the outputs, and develops a general model that describes the relationship.