Designing neural networks for modeling biological data: A statistical perspective

Cira Perna, Michele La Rocca
2013 Mathematical Biosciences and Engineering  
In this paper, we propose a strategy for the selection of the hidden layer size in feedforward neural network models. The procedure herein presented is based on comparison of different models in terms of their out of sample predictive ability, for a specified loss function. To overcome the problem of data snooping, we extend the scheme based on the use of the reality check with modifications apt to compare nested models. Some applications of the proposed procedure to simulated and real data
more » ... d and real data sets show that it allows to select parsimonious neural network models with the highest predictive accuracy. 2010 Mathematics Subject Classification. Primary: 62G08, 62H15; Secondary: 62F40, 92B20.
doi:10.3934/mbe.2014.11.331 pmid:24245721 fatcat:ncrwhrqcjrhfxjv6zctyfpjxqe