A Novel and Effective method for Early Identification of Cervical Cancer based on Gradient Boosting Classifier

In its initial stages, cervical cancer, which is the fourth greatest cause of death amongst women, shows no signs at all. At this time, cervical cancer can only be diagnosed using a limited number of methodologies that make use of machine learning methods. Certain methods, such as Hinselmann, Schiller, Cytology, and Biopsy, require the participation of a member of the medical staff, and cancer cannot be discovered until a particular stage has passed. In addition, many operations are prohibitively expensive in low-income countries. Using machine learning for the identification of cervical cancer is one potential approach to resolving this problem. This research study investigates the Gradient Boosting Classifier (GBC) for the diagnosis of cervical cancer and conducts trials with it. Kaggle's Cervical Cancer Data Set was utilized for this study's data collection. In the dataset about cervical cancer, there are a total of 32 risk factors, as well as four target parameters, namely Cytology, Hinselmann, Schiller, and Biopsy. After that, a comparison is made between GBC and other methods already in use, such as the Decision Tree technique and the Random Forest technique, and conclusions are produced regarding which technique is more effective at achieving the targets.