2024-03-29T02:05:50Z
https://meral.edu.mm/oai
oai:meral.edu.mm:recid/3308
2021-12-13T05:41:02Z
1582963302567:1597824273898
user-ucsy
Accuracy Comparison of Inlier Method and Random Row Method (Heart-disease)
Aye, Nan Saung Chan
Aye, Nilar
Heart disease is the leading cause of deathin the world over the past 10 years. Researchershave been using several data mining techniques tohelp health care professionals in the diagnosis ofheart disease patients. Decision Tree is one of thedata mining techniques used in the diagnosis ofheart disease showing considerable success. Kmeansclustering is one of the most popularclustering techniques; however initial centroidselection strongly affects its results.In this paper,heart-disease dataset is considered for study.Theimplemented system will be useful to find out thepatient’s level in the heart-diseases. This paperimplements integrating initial centroid selection ofthe k-means clustering such as inlier and randomrow methods with decision tree in the diagnosis ofheart disease patients. The result shows thatintegrating k-means clustering with decision treewith initial centroid selection could enhance theaccuracy in diagnosing heart disease patients. Italso shows that the inlier initial centroid selectionmethod could achieve higher accuracy thanrandom row initial centroid selection methods inthe diagnosis of heart disease patients.
2017-12-27
http://hdl.handle.net/20.500.12678/0000003308
https://meral.edu.mm/records/3308