Aim: The aim of this study is to classify Angina pectoris disease and compare the estimates of the methods by applying J48 and Random Forest methods, which are among the decision tree models, on the open access angina pectoris data set.
Materials and Methods: In the study, the data set named "Project Angina Data Set" was obtained from https://www.kaggle.com/snehal1409/predict-angina. In the data set, there are a total of 200 patients in whom angina pectoris was evaluated. Decision tree models J48 and Random Forest methods were used to classify angina pectoris disease.
Results: From the applied models, from the performance values obtained from the J48 method, the accuracy was 0.868, balanced accuracy 0.868, sensitivity 0.895, specificity 0.842, positive predictive value 0.85, negative predictive value 0.889 and F1-score 0.872. From the performance values obtained from the Random Forest method, the accuracy was 0.921, balanced accuracy 0.921, sensitivity 0.895, selectivity 0.947, positive predictive value 0.944, negative predictive value 0.9 and F1-score 0.919.
Conclusion: Considering the findings obtained from this study, it has been shown that the decision tree models used give successful predictions in the classification of angina pectoris disease.
Classification decision trees J48 Random Forest angina pectoris
Birincil Dil | İngilizce |
---|---|
Konular | Elektrik Mühendisliği |
Bölüm | Articles |
Yazarlar | |
Yayımlanma Tarihi | 31 Aralık 2020 |
Yayımlandığı Sayı | Yıl 2020 Cilt: 5 Sayı: 2 |