Abstract
Artificial Neural Networks (ANNs), which have been widely used to deal with classification problems, have been studied for decades. Different algorithms for synthesizing ANNs have been proposed as well. The number of neurons in an ANN usually affects the effciency of calculation in an ANN, either in the training phase or the recalling phase. That is, the fewer neurons used, the faster the calculationcan be performed. Furthermore, if the neurons are implemented by physical devices,the fewer number of neurons in an ANN reduces the implementation cost. In this paper, we propose a method to minimize the number of neurons used in an ANN that is built by using Voronoi diagrams while preserving its functionality. We conducted experiments on a set of benchmarks. The experimental results show that the resultant ANNs reduce the number of neurons by up to 94%, and by 37% onaverage.