Cervical Cancer Diagnosis Model Using Extreme Gradient Boosting and Bioinspired Firefly Optimization

Cervical cancer is frequently a deadly disease, common in females. However, early diagnosis of cervical can reduce the mortality rate and other associated complications. risk factors aid diagnosis. For better accuracy, we proposed study for using reduced feature set three ensemble-based classification techniques, i.e., extreme Gradient Boosting (XGBoost), AdaBoost, Random Forest (RF) along with Firefly algorithm optimization. Synthetic Minority Oversampling Technique (SMOTE) data sampling technique was used to alleviate imbalance problem. Risk Factors set, containing 32 risks factor four targets (Hinselmann, Schiller, Cytology, Biopsy), study. The are widely test cancer. effectiveness evaluated terms sensitivity, specificity, positive predictive accuracy (PPA), negative (NPA). Moreover, features selection achieve results number features. Experimental reveal significance model achieved highest outcome Hinselmann when compared diagnostic tests. Furthermore, reduction has enhanced outcomes. Additionally, performance models noticeable benchmark studies set.