Enhanced Prediction of Cervical Cancer Risk by combined Machine Learning Algorithms

Cervical Cancer constitutes a profound global health challenge, affecting a substantial number of women annually. If this cancer is not diagnosed and attended with hypervigilance, it can spread to other body parts, damage tissues, and typically deteriorate the immune system of the body. Deviating from the conventional methods of Cervical Cancer Prediction using Machine Learning (ML) Techniques, the proposed work Risk Prediction of Cervical Cancer focuses on the development and prediction of a model for early prediction of Cervical Cancer using combined ML techniques. The combined ML techniques proposed in the work includes - Support Vector Machine (SVM) and Gradient Boosting (GradBoost), Decision Tree (DT) and K-Nearest Neighbor (KNN), Random Forest (RF) and Extreme Gradient Boosting (XGBoost) techniques are employed to accurately anticipate cervical cancer diagnosing patients early and achieving better outcomes. The study's methodology involves dataset analysis, preprocessing, model training, and evaluation, culminating in the identification of Random Forest as an effective classifier. The proposed work outlines the necessity of combining ML techniques by achieving a significant result of Accuracy - 96%, Precision - 98%, Recall- 98% and F1-score - 98%. The proposed work underlines the necessity of collaboration between researchers, healthcare professionals, and governments to reduce cervical cancer worldwide and suggests future directions for improving prediction accuracy through the integration of new technologies and data enhancement.