Machine Learning Approaches for Disease Diagnosis Cervical Cancer Disease Prediction, A Case Study

Machine learning (ML) helps the healthcare sector to identify and forecast illnesses. Examining huge medical databases for hidden trends changes the diagnosis of diseases. Every year, hundreds of people die from cervical cancer. Correct and accurate diagnosis determines the quality of life and treatment efficacy. Reduced cervical cancer death and incidence depends on improved accuracy and faster detection. Predicting cervical cancer with ML might help to improve healthcare treatment and lower expenses. Conventional diagnostic techniques' low sensitivity and specificity could cause delays or misdiagnoses. This case study shows how machine learning may improve cervical cancer detection and lower errors. This paper describes a two-part experiment based on the Cervical Cancer Disease Dataset kept in the UCI-ML Repository. Sequential feature selection lets several ML methods be evaluated on the dataset. Various ensemble learning approaches ensemble helped to enhance the first phase of the experiment. Consequently, we were able to attain better prediction results using the same dataset: an F1-score of 96.97%, an accuracy of 95.45%, a precision of 94.12%, a sensitivity of 100%, and a specificity of 92.85%. Lastly, ROC curves with AUC values of 0.982 were given in the empirical investigation.