This paper studies one of intelligent data processing methods: using association rules for data analysis. The method of association rule obtaining what was initially developed to analyse consumer’s basket has turned to be a good tool for other tasks too. The method helps search and find regularities of the form X Y in different kinds of data. Nowadays this method is widely applied in the tasks of large scale database processing and analysing. As a result, methods of association rule construction occupy their place among the basic methods of intelligent data processing. The paper consists of two parts: theoretical and experimental. The theoretical part examines the mathematical aspects of association rule construction in detail and describes basic concepts and algorithm application possibilities. The experimental part presents implementation results and analysis of experiments. Conclusions have been drawn concerning the efficiency of association rules’ application in search of regularities. Even though the association rules mining method is among the fundamental data processing methods, in Latvia this method is not widely used, therefore, the article under consideration reveals the potential possibilities of the association rule mining in the analysis of statistical data.
Multi-task learning for intelligent data processing in granular computing context