Improving Breast Cancer Diagnosis: Insights From Machine Learning Models

Data mining is a technique used in the healthcare industry to analyze vast amounts of information for early illness diagnosis and prognosis. When it comes to its early prediction stages of breast cancer, many different classification algorithms play an essential role. This study proposed a pragmatic method to evaluate the effectiveness of four different classification algorithms, namely Decision Tree, Naïve Bayes, K-nearest neighbor and Support Vector Machines, in accordance to forecast breast cancer. These algorithms were examined for their ability to diagnose and predict the early stages of illness. The classifier is strengthened by using exploratory data analysis and the boosting method in this study. Additionally, the problem of imbalanced classes is resolved during the classification phase. In this particular instance, cross-validation techniques were utilized to investigate the behavior of each classifier concerning the accuracy, true positive rate, and false positive rate. This study attempted for improvement of the accuracy level in the prediction of breast cancer employing through the aforementioned machine-learning methods.