Computational performance of hypercube reduction methods for multidimensional data of analytical OLAP system

The paper investigates mathematical methods of decomposition (reduction) of large hypercubes of multidimensional data of analytical OLAP-systems into subcube components. The criterion for reducing the computational complexity of solving these problems by decompositional methods of exponential and polynomial-logarithmic degrees of complexity compared with traditional methods for analyzing large amounts of information accumulated in hypercubes of multidimensional OLAP data is proved. For reduction methods for analyzing OLAP cubes of a logarithmic degree of complexity, a criterion is established for increasing computational complexity in comparison with non-reduction methods. An exact upper bound for the change in the complexity of decomposition data analysis methods for varying the main parameters of the hypercube is obtained.