摘要：An issue of building a tool orientated to the support of selection process of decision support system (DSS) for the sector of small and medium enterprises is presented in the article.A group method of data handling (GMDH) whose application allows to objectives a search process of required DSS,with assumed costs and other existing resource limitations,is proposed for this purpose.The article is devoted to a problem of GMDH adjustment to solving the tasks related to the selection of a required system from among the information systems of DSS class,according to the criteria determined by a future user of this system.The GMDH takes into account among other things such assumptions as: a precise description of dependences between input and output data in a specified time horizon,independence of the user's knowledge (the values of observed data from the past registered in the base are the input data) and minimization of modelling errors (the selection of solutions takes place by means of a selected evaluation rule of quality estimation).
关键词：Decision support system (DSS);multi-criteria decision making model;group method of data handling (GMDH);information system of GMDH class;neuronal networks;genetic algorithms;root mean squared error of approximation;Iwachnienko polynomial;polynomial decisi
