The detection polinomial algorithm decision NP tasks MDX prediction IN mode OF parallel computing

Abstract

The task in this work is and is forecasting in MDX economic area with the use of analytical mathematical models built on multidimensional statistical data. Set of models forms a knowledge base to generate full Combinatorial set of forecasting scripts. Each script realizes iterative procedure of definition regular stable forecast point. There are the appraisals of the amount of the script depending on the dimension of economic space. In theory it’s disclosed that MDX prediction is a NP task with the nondeterministic algorithm. In practice it`s revealed that time of serial solution of the task forecasting exponentially depends on the dimensions of space. Parallel decision task of forecasting is fulfilled polynomials that is confirmed experimentally for different types of forecasting, for example, population of Germany, GDP and other settings MDX economic space of its statistics.