OPTIMIZED PREDICTIVE MODELLING FOR CERVICAL CANCER DETECTION USING MACHINE LEARNING ALGORITHMS

Abstract

Cervical cancer is one of the most prevalent and life-threatening diseases affecting women worldwide. Early detection through data-driven approaches plays a crucial role in reducing mortality and improving treatment outcomes. This study utilizes a real-world cervical cancer dataset containing demographic, behavioral, and medical attributes such as age, sexual behavior , smoking habits, contraceptive use, and history of sexually transmitted diseases (STDs). The objective is to develop a machine learning-based predictive model to assess cervical cancer risk and enable early screening and timely intervention. Various preprocessing techniques, including handling missing values and applying the Synthetic Minority Over-sampling Technique (SMOTE) for class imbalance, were used to enhance model performance. The trained Random Forest model achieved high accuracy and demonstrated strong predictive capability, highlighting the potential of machine learning in supporting preventive healthcare and early detection of cervical cancer.