Jing Chang and Herbert K. H. Lee
11/03/2012 10:50 PM
Applied Mathematics & Statistics
A graphical tool for choosing the number of nodes for a neural network is introduced. The idea here is to t the neural network with a range of numbers of nodes at first, and then generate a jump plot using a transformation of the mean square errors of the resulting residuals. A theorem is proven to show that the jump plot will select several candidate numbers of nodes among which one is the true number of nodes. Then a single node only test, which has been theoretically justified, will be used to rule out erroneous candidates. The method has a sound theoretical background, yields good results on simulated datasets, and shows wide applicability to datasets from real research.
