Optimization of the Heart Disease Prediction Model using Adaboost based on the Decision Tree Sri Mulyati (*), Endar Nirmala, Aries Saifudin, Yulianti, Irpan Kusyadi, Teti Desyani
Universitas Pamulang
Jl. Raya Puspitek No. 46 Buaran, Serpong, Tangerang Selatan, Banten, Indonesia, 15417
*dosen00391[at]unpam.ac.id
Abstract
Coronary heart disease has become one of the leading causes of death worldwide, and coronary heart disease mortality has increased from year to year. Early detection of coronary heart disease and by performing manual health risk assessments. Has less comprehensive data about relevant risk factors and has limited ability to predict the risk of developing coronary heart disease in Indonesia. Opportunities for machine learning appear to be enormous in computer-assisted Parkinson^s classification and can reduce inevitable misdiagnosis and variability in healthcare, provide guidance (in the absence of a specialist) and quick decision-makers. Many models have been introduced in the literature to diagnose coronary heart disease using various machine learning algorithms, but have not produced a perfect performance with a variety of training data. The use of the Decision Tree has been used successfully for medical prediction and reliable decision-making techniques. The prediction results using the Decision Tree are not perfect because they still contain prediction errors. So that in this study proposed the application of AdaBoost to reduce prediction errors. In general, the Boosting algorithm is better than Bagging, but not evenly good. AdaBoost could theoretically be used significantly to reduce the errors of some learning algorithms which consistently resulted in better classifier performance. Based on the experiments conducted, the application of AdaBoost can improve the performance of the Decision Tree in predicting coronary heart disease.
