Introduction to Knowledge Discovery in Medical Databases and Use of Reliability Analysis in Data Mining
Elena Zaitseva, Miroslav Kvassay, Vitaly Levashenko, Jozef Kostolny
Citation: Proceedings of the 2015 Federated Conference on Computer Science and Information Systems, M. Ganzha, L. Maciaszek, M. Paprzycki (eds). ACSIS, Vol. 5, pages 311–320 (2015)
Abstract. Data mining (DM) is a collection of algorithms that are used to find some novel, useful and interesting knowledge in databases. DM algorithms are based on applied fields of mathematics and informatics, such as mathematical statistics, probability theory, information theory, neural networks. Some methods of these fields can be used to find hidden relation between data, what can be used to create models that predict some behavior or describe some common properties of analyzed objects. In this paper, we combine methods of DM with tools of reliability analysis to investigate importance of individual database attributes. Results of such investigation can be used in database optimization because it allows identifying attributes that are not important for purposes for which the database is used. Our approach is based on some coincidence between the key terms of DM and reliability analysis.