A Hybrid Model for Detection of Cervical Cancer using Single Target variable Applying Machine Learning Approaches

Department of Computer Science and Engineering Birla Institute of Technology Mesra, India
Department of Computer Science and Engineering Birla Institute of Technology Mesra, India
Department of Management Studies Haldia Institute of Management Haldia, India
Department of Management Birla Institute of Technology Mesra, India

This email address is being protected from spambots. You need JavaScript enabled to view it.
This email address is being protected from spambots. You need JavaScript enabled to view it.
This email address is being protected from spambots. You need JavaScript enabled to view it.
This email address is being protected from spambots. You need JavaScript enabled to view it.

Abstract

There are a number of cancer diseases which attack different organs of a human body. More or less all types of cancers are fatal in nature. For women, cervical cancer appears as a very painful, suffering and life threatening disease. Early detection is the key way to reduce its fatal and painful effects. There are many methods available in health care industries to diagnose cervical cancer. The doctors or medical professionals are the ones who perform the analysis of findings from medical reports and determine their course of actions. Traditional medical examinations and treatment of diseases have been changed to some extent with the advent of computer science and technologies. Medical services adopt various tools and techniques for sophistication and perfection of treatment processes. Artificial intelligence and machine learning nowadays are in high choice for pattern recognition and disease analysis. It is proved to be very effective and useful in identification of diseases analyzing relevant data. This study presents a hybrid model using cytology target variable applying a combination of machine learning algorithms namely Extreme Gradient Boosting (XGB), Synthetic Minority Over sampling Technique (SMOTE) and Chi-Square statistical method. The model proceeds with fold by fold approach. Many researchers have already worked with this field using machine learning and many of them represented very good models. The proposed study works with efficient selection of features needed for classification purposes and after running the sophisticated algorithms the study succeeds to present higher accuracy and better performances.