Claim Missing Document
Check
Articles

Found 1 Documents
Search
Journal : Jurnal Sisfotek Global

The Utilization of Decision Tree Algorithm In Order to Predict Heart Disease Mia Mia; Anis Fitri Nur Masruriyah; Adi Rizky Pratama
JURNAL SISFOTEK GLOBAL Vol 12, No 2 (2022): JURNAL SISFOTEK GLOBAL
Publisher : Institut Teknologi dan Bisnis Bina Sarana Global

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.38101/sisfotek.v12i2.551

Abstract

The data on heart disease patients obtained from the Ministry of Health of the Republic of Indonesia in 2020 explains that heart disease has increased every year and ranks as the highest cause of death in Indonesia, especially at productive ages. If people with heart disease are not treated properly, then in their effective period a patient can experience death more quickly. Thus, a predictive model that is able to help medical personnel solve health problems is built. This study employed the Random Forest and Decision Tree algorithm classification process by processing cardiac patient data to create a predictive model and based on the data obtained, showing that the data on heart disease was not balanced. Thus, to overcome the imbalance, an oversampling technique was carried out using ADASYN and SMOTE. This study proved that the performance of the ADASYN and SMOTE oversampling techniques on the C45 algorithm and the Random Forest Classifier had a significant effect on the prediction results. The usage of oversampling techniques to analyze data aims to handle unbalanced datasets, and the confusion matrix is used for testing Precision, Recall, and F1-SCORE, as well as Accuracy. Based on the results of research that has been carried out with the K-Fold 10 testing technique and oversampling technique, SMOTE + RF is one of the best oversampling techniques which has a greater Accuracy of 93.58% compared to Random Forest without SMOTE of 90.51% and SMOTE + ADASYN of 93.55%. The application of the SMOTE technique was proven to be able to overcome the problem of data imbalance and get better classification results than the application of the ADASYN technique.