Scientific publications of the Ivan Franko National University of Lviv

Visnyk of the Lviv University. Series Applied Mathematics and Computer Science

Additional information
Year 2003
Volume 6
Authors Nikolski I., Shcherbyna Yu., Yakymechko R.
Name of the article Decision trees and their applications for diagnosis in medicine
Abstract The technology of decision tree applications is described in detail in this article. The types of decision-making problems are formulated, the precise definition of decision trees and decision-making systems are given. ID3 algorithm of decision tree building is described and its application is illustrated in two model problems. Methods of knowledge-based system based on the use of decision trees with the application of ID3 algorithm were worked out on the real information collected at giving the diagnosis of a cardio disease. The diagnosis was given by analyzing fifteen indicators (see table 3). The attribute of making decisions in table 3 is attribute KHKS where meaning 1 corresponds to disease presence, 0 – to its absence. This table contains results of 3532 examined patients. In order to check the quality of the received results the data where divided into two parts where the first was used to decision tree building (and corresponding rule base), the other one for checking the prognosis quality. For this, all lines from the second part of the table where used as input examples of the built rule base and the known diagnosis was compared to the received application of the built rule base. The practical purpose of the carried out research was the decision tree building and its fundamental rule sets which wood give an opportunity to automate the process of giving diagnosis and thus create a prototype of the expert system.
Language Ukrainian
DOC format Nikolski I., Shcherbyna Yu., Yakymechko R. 
small logo
©2003-2012 Lviv University | Contacts