Pre Screening of Cervical Cancer Through Gradient Boosting Ensemble Learning Method

Abstract

In recent years, cervical cancer is one of the most common diseases which occur in any woman regardless of any age. This is the deadliest disease since there were no symptoms shown till it is diagnosed to be the last stage. For women at a certain age, it is better to have a proper screening for cervical cancer. In most underdeveloped nations, it is very difficult to have frequent scanning for cervical cancer. Data Mining and machine learning methodologies help widely in finding the important causes for cervical cancer. The proposed work describes a multi-class classification approach is implemented for the dataset using Support Vector Machine (SVM) and the perception learning method. It is known that most classification algorithms are designed for solving binary classification problems. From a heuristic approach, the problem is addressed as a multiclass classification problem. A Gradient Boosting Machine (GBM) is also used in implementation in order to increase the classifier accuracy. The proposed model is evaluated in terms of accuracy, sensitivity and found that this model works well in identifying the risk factors of cervical cancer.