STUDY OF OPPORTUNITIES OF GENETIC ALGORITHMS IN THE CLUSTERING TASK USERS ON THE INTERNET

Abstract

This paper proposes an approach for solving the clustering network users Internet. A genetic algorithm to solve the problem and the appropriate software software for its practical application. To take into account the specifics of the problem and improve effectiveness of the genetic algorithm were applied heterogeneous chromosome. In accordance with the This included substantial modifications to the procedures for the classical crossover and mutation. The developed algorithm has been tested for performance in comparison with the threshold algorithm sorting and shown its significant advantage for this indicator. Further studies will be linked with the development of the intellectual approach to this problem and the use of fuzzy sets to overcome the noise input.