The work is devoted to the problem of price motion prediction on the stock market. An interaction of a big number of people involved in the investment process is considered from the chaos and self-organization theory point of view. The method of short-term predictions building is developed based on nonlinear and neuro-science conceptions. A detailed analysis of practical results gained on operation testing of the prediction system on the Russian stock market allows to make a conclusions about adaptability of investigating method and also planning main prospects of this method development.

Mathematical modelling in actual problems of science and technics