Cervical Cancer Behavior Risk Prediction Using Machine Learning

Abstract

Cervical cancer is a serious public health concern that affects women all over the world. Early risk prediction of cervical cancer can play an essential role in prevention by boosting public awareness of this disease, because it is a fatal disease. Both healthcare professionals and persons at risk can benefit from early prediction utilizing a Machine Learning (ML) model. Using a dataset from the UCI ML repository, eleven supervised machine learning algorithms are used to predict early risks of cervical cancer in this work. Accuracy, precision, F1, recall, and ROC_AUC are among the performance metrics used to predict the early risks of cervical cancer using machine learning models. Finally, a comparison analysis reveals that using the Multi-Layer Perceptron (MLP) method with default hyperparameters, this study achieved 93.33% prediction accuracy. Decision Tree Classifier (DTC), Random Forest Classifier (RFC), K- Nearest Neighbors (KNN), Support Vector Machine (SVM), and Multi- Layer Perceptron (MLP) all showed accuracy of 93.33% when using the hyperparameter tuning strategy.