MULTIPLE-CRITERIA COMPARATIVE ANALYSIS OF DATA MINING METHODS FOR USING IN DBMS

The most popular methods of Data Mining are investigated: classical methods, neural networks, visualization methods, decision trees, polynomial neural networks, k-neighboring neighbors. These methods are used to conduct statistics and search for regularities in large volumes of data. To construct the rating of the considered methods a quantitative multiple-criteria method the analytic hierarchy process in absolute measurements was chosen. The resulting rating can be used to select the Data Mining method for a specific task.