MATHEMATICAL CONCEPTS OVERVIEW OF CLUSTER ANALYSIS, SVM AND NEURAL NETWORKS IN MASS SERVICE NETWORKS

Abstract

Nowadays, mass service networks are researched by using a variety of methods. However, the results that have been achieved this far are intermediate ones, especially, in case of multiservice telecommunication networks. The present paper provides an overview of cluster analysis, SVM and neural networks algorithms. The comparative analysis of mentioned models’ prediction results was conducted and it was practically showed that that neural networks can provide a powerful tool for the telecommunications industry because the Neural Networks are nonlinear dynamic systems able to offer a solution almost anywhere where classic methods have failed. Neural Networks have certainly become one of the key technologies nowadays, though a critical analysis reveals some drawbacks.