Abstract
A new two-dimensional feed-forward functionally expanded neural network (2D FFENN) used to produce surface models in two dimensions is presented. New nonlinear multilevel surface basis functions are proposed for the network's functional expansion. A network optimization technique based on an iterative function selection strategy is also described. Comparative simulation results for surface mappings generated by the 2D FFENN, multilevel 2D FFENN, multilayered perceptron (MLP), and radial basis function (RBF) architectures are presented.
Highlights
One of the main properties of feed-forward neural networks is that of learning an input-output mapping from a set of examples characterizing a real system
The pruning process is stopped at the stage when a pruned network structure is found to be incapable of reducing the output maximum surface level error (MSLE) or mean square error (MSE) on the training set to the desired level
In order to illustrate the modeling capability of the 2D FFENN structure and the effectiveness of the pruning strategy, we produce a model for the smooth continuous surface described by (19)
Summary
One of the main properties of feed-forward neural networks is that of learning an input-output mapping from a set of examples characterizing a real system. Two well-known feed-forward artificial neural networks are the multilayered perceptron (MLP) and radial basis function (RBF). The design of a new single hidden layer, linear in the parameters, two-dimensional feed-forward functionally expanded neural network surface modeler (2D FFENN) is presented. A 1D FFENN has been successfully applied to time series prediction [8, 9] and cochannel interference [10] The aim of this new design is to explore the modeling capabilities of such a feed-forward network in two dimensions.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
