The main purpose of this paper is to compare the support vector machine (SVM) developed by Vapnik with other techniques such as Backpropagation and Radial Basis Function (RBF) Net...
The performance characteristics of five variants of the Hopfield network are examined. Two performance metrics are used: memory capacity, and a measure of the size of basins of att...
In this paper we aim to train deep neural networks for rapid visual recognition. The task is highly challenging, largely due to the lack of a meaningful regularizer on the functio...
Abstract. This paper examines the generalization capability in learning multiple temporal patterns by the recurrent neural network with parametric bias (RNNPB). Our simulation expe...
: The initialisation of a neural network implementation of Sammon's mapping, either randomly or based on the principal components (PCs) of the sample covariance matrix, is exp...
Boaz Lerner, Hugo Guterman, Mayer Aladjem, Its'hak...