Decision Tree Algorithms for Predictive Modeling in Breast Cancer Treatment
Huda Kutrani, Saria Eltalhi · 2022 IEEE 2nd International Maghreb Meeting of the Conference on Sciences and Techniques of Automatic Control and Computer Engineering (MI-STA) · 2022
Successful treatment of breast cancer increases a high chance of survival among women. Machine Learning could support the discovery of knowledge and important patterns from medical data by a good predictive model. The study aims to identify the effective and predictive modeling for the primary treatment of breast cancer after diagnosis. The data-set was collected from patients records who have malignant from the Benghazi Medical Center. Decision Tree algorithms J48, CART, and Random Forest were used to build three models by WEKA software. The data-set was divided into two subsets training and testing. Accuracy of Prediction, Sensitivity test, Specificity test, Area under curve, Kappa statistics, and Mean Absolute Error was used to compare models’ performance. The study results showed that the Random Forest was the better model for "training data" and "test data" compared to J48 and CART models. Also, the Random Forest model could provide good information and is able to recognize the important patterns. This study concluded that the Random Forest might a good model for detecting breast cancer treatment.