APPLICATION OF GENETIC ALGORITHMS TO INCREASE AN OVERALL PERFORMANCE OF NEURAL NETWORKS IN THE DOMAIN OF DATABASE STRUCTURES SYNTHESIS

Abstract

In the given paper a method is described to augment an artificial neural network with corresponding genetic optimization algorithm inside the unified algorithmic framework. Application of this approach is demonstrated for the synthesis of optimum logical structure of a distributed database that has a great impact on an effective design of modern distributed information infrastructure. A neural network algorithm gives local-optimum solutions of that extreme task, while the function of the genetic algorithm within the framework of this problem is the selection among the received local-optimum solutions a global-optimum, satisfying to the given constraints of the problem.