Design and evaluation of neural classifiers
L. Hintz-Madsen, M.W. Pedersen, Lars Kai Hansen, Jan Otto Larsen · 2002
In this paper we propose a method for the design of feedforward neural classifiers based on regularization and adaptive architectures. Using a penalized maximum likelihood scheme we derive a modified form of the entropy error measure and an algebraic estimate of the test error. In conjunction with optimal brain damage pruning the test error estimate is used to optimize the network architecture. The scheme is evaluated on an artificial and a real world problem.