№3 (2018)

Samigulina G.A., Masimkanova Zh. A.

The article is devoted to the actual problem of application of bioinspired intellectual approaches for preliminary data processing and solving the problem of prediction in the oil and gas industry. The aim of the research is to construct an optimal immune network model based on modified algorithms of ant colonies and artificial immune systems. The selection of informative descriptors is carried out based on a multi-algorithm approach, which allows to use of several modified algorithms. The ontological models of considered algorithms have been proposed for the systematization of used methods and the facilitation of writing of software. Modified algorithms of ant colony differ by the presence of additional variables, various constraints, topology, etc. Such parameters as the size of populations of agents and the amount of iterations have great importance.
Keywords: modified algorithms of ant colonies, immune network modeling, the selection of informative descriptors.