A Comparative Analysis of Tree-based Machine Learning Algorithms for Breast Cancer Detection
Fiddin Yusfida A’la, Adhistya Erna Permanasari, Noor Akhmad Setiawan · 2019
Breast cancer prediction plays an essential role in medical decision making. This paper presents various machine learning algorithm based on three categories of decision tree learners to predict the presence of breast cancer based on routine blood analysis. Prediction, the presence of breast cancer, is performed in two scheme: non-feature selection and feature selection in the pre-processing stage. Machine learning that we used in this study is a basic decision tree, random forest, and gradient boosting. The performance of these classifiers is evaluated concerning sensitivity and specificity. The result shows that gradient boosting reached 85% of sensitivity and value of specificity is 80%. From this result, we can conclude that the combination of feature selection in the pre-processing step and gradient boosting classifier performs better than other schemas.