Improving Accuracy: Comparative Analysis of Machine LearningModels for Prostate Cancer Prediction

Among the different types of cancer affecting men is prostate cancer, which ranks second in mortality after lung cancer, a worrying reality. Nowadays, Machine Learning (ML) models have contributed to different areas, being their contribution to the medical field one of the most outstanding. This study aims to compare the accuracy of ML models in the prediction of prostate cancer. Gradient Boosting (GB), Random Forest (RF), Decision Tree (DT) and Adaptive Boosting (AdaBoost) models were analyzed. In addition, DT, RF, K-Nearest Neighbors (KNN), Gaussian Naive Bayes (GNB) and Logistic regression (LR) models were used to identify the base model for algorithm optimization. The study was divided into several stages, such as the description of the models and the analysis of the data set, among others. On the other hand, the metrics of sensitivity, precision, specificity, accuracy, and F1 count were used to contrast the algorithms. The training results positioned the GB algorithm as the most accurate algorithm for prostate cancer detection with 83.33% accuracy, 98.02% precision and 95.24% sensitivity.