Separation of the informative attributes of the strains plague microbe for forecasting epizootic process of the plague

G.A.Samigulina, S.V.Chebeiko

Institute of Informatics and Control Problems of the Republic Kazakhstan,

Pushkin Street, 125, office 102, Almaty, 480100, Kazakhstan

e-mail: galinasamigulina@mail.ru

 

љљљљљљљљљљљ

The business- economic activity of the man plays the huge role in spreading of the especially dangerous infection, to which to concern the plague, from its primary centers. Increase of the density of the population, developing of territories of the natural center plague, the development of the international, economic and tourist connections promotes spreading of the plague in the world.љ At present the complex mechanisms of interaction between the members of the plague epizootic triad: the agent of the plague- the plague microbe, the host - the rodent (large gerbil) and the vector- the flea are not enough investigated. For realization of the authentic forecast of the epizootic plague and intensity spreading of the infection it is necessary to development of new technology.

Now on the border of the theory of control and the artificial intelligence the new area of researches- the intellectual control is developed impetuous. The problems by solving with the help of the intellectual control consist of the number of the tasks. One of the most widespread and actual tasks during many years is the task of forecasting.

For creation of the new technologies of forecasting of the epizootic plague the perspective approach Artificial Immune Systems (AIS) is used. AIS based on the biological principles of processing of the information by the molecules of the protein.

Given paper is continuation of past work, in which on the basis of the approach AIS the task of the complex estimation of the conditions the plague triad, prediction of the beginning epizootic plague and intensity of the spreading infection is solved. However, the processing of the multidimensional data set on the base of AIS became complicated of the number serious problems, which influence on the reliability of the forecast. There are errors of the energy estimations AIS because of the problem incompleteness of the data, its correlation and the errors of measurement.

The task of decreasing errors of the energy estimation at forecasting epizootic of the plague on the basis of the approach AIS and separation of the informative attributes is decided on the basis of the methods factor analysis and to development of the varimax approach.