Overlearning in Neural Networks

When an iterative training algorithm is run, over-fitting occurs when the algorithm is run for too long, and the network is too complex for the problem or the available quantity of data. See STATISTICA Neural Networks.