AUTHORS: Irina Astachova, Stanislav Ushakov, Andrei Selemenev, Juliya Hitskova

Download as PDF

ABSTRACT: Ecological prognoises sets the identification task, which is - to find the capacity of pollution sources based on the available experimental data. This problem is an inverse problem, for the solution of which the method of symbolic regression is considered. Some authors solved a similar problem with the help of neural networks. In this paper the distributed artificial immune system is used as an algorithm for the problem solving. The artificial immune system (AIS) is a model that allows solving various problems of identification, its concept was borrowed from biology. The solution is sought using a distributed version of the artificial immune system, which is implemented through a network. This distributed network can operate in any heterogeneous environment, which is achieved through the use of cross-platform Python programming language. AIS demonstrates the ability to restore the original function in the problem of identification. The obtained solution for the test data is represented by the graph. This language has been chosen for the following reasons: cross-platform reason: during the creation of an application, running in a heterogeneous computing environment, the cross-platform factor is very important; the speed of development reason: Python is oriented on the increase of developer productivity and code readability; the standard library reason: includes a large number of useful functions and classes for various tasks such as networking, multithreading, process management.

KEYWORDS: Artificial Immune Systems (AIS), symbolic regression, distributed calculations, identification problem in ecology

REFERENCES:

[1] A.N. Vasilev, D.A. Tarhov Parametric neural network models of building of regularization of identify problems solution in ecology. Contemporary information technologies and ITeducation 1 (2014).

[2] D. Dasgupta. Artificial Immune Systems and their Application: collection of papers. Moscow.: FIZMATLIT (2006).

[3]. D. Dasgupta, S. Yua, F. Nino. Recent Advances in Artificial Immune Systems: Models and Applications. Appl. Soft Computing 11 (2011).

[4] F. Freschi, M. Repetto. Multiobjective optimisation by a modified artificial immune system, Artificial Immune Systems (2005), 248– 261.

[5] N.X. Hoai, R.I. McKay, D. Essam. Solving the symbolic regression problem with tree-adjunct grammar guided genetic programming: the comparative results. Evolutionary Computation, CEC '02. Proceedings of the 2002 Congress 2 (200) 1326-1331.

[6] I. E. Hunt, D. E. Cooke. Learning using an artificial immune system. J. Network Comp. Appl. 19 (1996) 189-212.

[7] A.A. Ishiguro, Y. Watanabe, T. Kondo. Robot with a decentralized consensus-making mechanism based on the immune system. Proceedings of ISADS (1997) 231-237.

[8] C.G. Johnson. Artificial immune systems programming for symbolic regression. Genetic Programming: 6th European Conference (2003) 345–353.

[9] J.O. Kephart. A biologically inspired immune system for computers. Proceedings of Artificial Life IV: The Fourth International Workshop on the Synthesis and Simulation of Living Systems (1994) 130–139.

[10] R.A. Schollmeier. Definition of peer-to-peer networking for the classification of peer-to-peer architectures and applications. Proceedings of the First International Conference on Peer-to-Peer Computing, IEEE (2001) 101-102.

[11] E. Hart, J. Timmis. Application areas of AIS: The past, the present and the future. Appl. Soft Computing 8 (2008) 191–201.

[12] M. Bardeen. Survey of methods to prevent premature convergence in evolutionary algorithms. Workshop of Natural Computing, J. Chilenas de Computacion (2013) 13–15.

[13] D. Barkai. Peer-to-Peer Computing. Santa Clara: Intel Press (2002).

[14] K. Bennett, M.C. Ferris, Y.E. Ioannidis. A genetic algorithm for database query optimization. Proceedings of the fourth International Conference on Genetic Algorithms (1991) 400-407.

[15] H. Bersini H. The endogenous double plasticity of the immune network and the inspiration to be drawn for engineering artifacts. Artificial Immune Systems and Their Applications (1999) 22-44.

[16] I.F. Astakhova, S.A. Ushakov, Ju.V. Hitskova. Model and algorithm of an artificial immune systems for the recognition of single symbols and their comparison with existing methods. WSEAS Trans. on Information Science and Applications 13 (2016) 38-45.

[17]J.R.Koza. Genetic Programming. Cambridge: MIT Press (1998). — 220 p.

[18] W.B. Langdon , R. Poli.. Foundations of Genetic Programming Haidelberg: SpringerVerlag (2002).