V. E. Bolnokin - Dr.Sc. (Eng.), Professor, The federal state unitary enterprise Research and experimental institute of automobile electronics and an electric equipment. E-mail: vitalybolnokin@yandex.ru, d.i.mutin@mail.ru
D. I. Mutin - Ph.D. (Eng.), Institute mashinovedenia the Russian Academy of Sciences. E-mail: d.i.mutin@mail.ru
E. I. Mutina - Ph.D. (Eng.), Assosiate Professor, Federal state budgetary educational institution of the higher vocational training the Moscow state technical university «Stankin». E-mail: e.mutina@mail.ru

The article describes the way to solve the problem of revealing hidden relationships in difficult structured data having implicit character, using data mining in particular - decision trees, the effectiveness of which is illustrated by a specific example. As an example the automated distributed medical system model health care facility, as well as decision support system, is a system of diagnosis and differential diagnosis. Implicit tasks medicine were ideal field for the application of data mining. In this context, proposes the creation of a decision support system based on the method of decision trees. Decision support systems are complex hardware and software systems are intended to assist decision-makers in the management of complex objects and processes of different nature. Information on the basis of which a decision must be accurate, complete, consistent and adequate. The organization decision support system used data warehouse, which act as pretreatment and storage on the basis of information from the automated control system of the organization and other sources. Approach to the search and retrieval of data, which is based on the idea of the existence of relationship between the frequency and the degree of aggregation queries data that queries operate is called On-line Analytical Processing OLAP. The basic concept is to OLAP multidimensional cube data, which are stored in the cells analyzed (numerical) data. The use of data mining, presented with the help of OLAP systems in the form of real or virtual information hypercube, in many cases, is more effective, and most importantly - much more closely integrated into a single information-analytical system. Accordingly, it is advisable to use a hybrid method of analysis of data, constructed on the joint application of the method of decision trees, and OLAP-analysis.