A COMPARISON OF HEART ABNORMALITIES DETECTION ON ECG USING KNN AND DECISION TREE

STIVEN HAMONANGAN SINURAT, CALVIN TIOPAN , MAGGIE AMELIA FONG (2021) A COMPARISON OF HEART ABNORMALITIES DETECTION ON ECG USING KNN AND DECISION TREE , SKRIPSI, UNIVERSITAS PRIMA INDONESIA

ABSTRAK

In life, the heart is always required to always be in good condition because the heart serves to pump blood that carries nutrients throughout the body. Impaired heart function can be fatal to human health, even some heart disorders can lead to death. To be able to detect the presence of abnormalities or disorders of the heart, it must be known in advance the working rhythm or signal pattern of the heart itself. In this study the algorithm we used is KNN, and decision tree of the three algorithms will be searched for the best results so that the initial diagnose error can be minimized. K-NN gives the best results accuracy of 97.373% and the decision tree gives the best results an accuracy of 95.87%. Early treatment or detection is important in order to reduce the risk of death that can be caused. Early detection of cardiovascular disease can be done with ECG wave analysis, in which case Artificial Intelligence can help early detection more easily and quickly. Implementing machine learning can help this process to be more efficient. Like this project, machine learning can classify abnormal ECGs with maximum accuracy 97.373%.

PROSIDING