This work studies some of the approximating properties of feedforward neural networks as a function of the number of nodes. Two cases are considered: sigmoidal and radial basis function networks. Bounds for the approximation error are given. The methods through which we arrive at the bounds are constructive. The error studied is the L1 or sup error.

Multilayer perceptrons and radial basis function networks are used most often in classification tasks, even though the two neural networks have different performance in classification tasks depending on the available training data sets. This paper shows the accuracy change in classification of the two neural networks when training data set size changes. Experiments were run with four data sets ...

A study is presented to compare the performance of multilayer perceptrons, radial basis function networks, and probabilistic neural networks for classification. In many classification problems, probabilistic neural networks have outperformed other neural classifiers. Unfortunately, with this kind of networks, the number of required operations to classify one pattern directly depends on the numb...

The typical system identi cation procedure requires powerful and versatile software means. In this paper we describe and exemplify the use of a prototype identi cation software tool, applicable for the rather broad class of multi input single output model structures with regressors that are formed by delayed inand outputs. Interesting special instances of this model structure category include, ...

A Bayesian-based methodology is presented which leads to a data analysis system based around a committee of radial-basis function (RBF) networks. We show that this approach enables estimatation of the uncertainty associated with system outputs. Systems with diiering numbers of internal degrees of freedom (weights) may hence be compared using training data only. Feedforward neural networks have ...

This paper proposes a framework for constructing and training a radial basis function (RBF) neural network. For this purpose, a sequential learning algorithm is presented to adapt the structure of the network, in which it is possible to create a new hidden unit and also to detect and remove inactive units. The structure of the Gaussian functions is modi"ed using a pseudoGaussian function (PG) i...

The goal of knowledge transfer is to take advantage of previous training experience to solve related but new tasks. This paper tackles the issue of transfer of knowledge between radial basis function neural networks. We present some preliminary work illustrating how a neural network trained on one task (the source) can be used to assist in the synthesis of a new but similar task (the target).

The behavior of Radial Basis Function (RBF) Networks greatly depends on how the center points of the basis functions are selected. In this work we investigate the use of instance reduction techniques, originally developed to reduce the storage requirements of instance based learners, for this purpose. Five Instance-Based Reduction Techniques were used to determine the set of center points, and ...

In this paper, we investigated an ensemble neural network for the prediction of oil prices. Daily data from 1999 to 2012 were used to predict the West Taxes, Intermediate. Data were separated into four phases of training and testing using different percentages and obtained seven sub-datasets after implementing different attribute selection algorithms. We used three types of neural networks: Fee...

A Neural Network (NN) using Normalised Radial Basis Functions (NRBF) is used for encoding the sequence of positions forming the path of an autonomous wheelchair. The network operates by continuously producing the next position for the wheelchair. As the path passes several times over the same point, additional phase information is added to the position information. This avoids the aliasing prob...