Colorectal Cancer Prediction Using Machine Learning and Neutrosophic MCDM Methodology: A Case Study

Juan Viteri Rodríguez^1,*, Julio Rea Martínez², Freddy F. Jumbo Salazar³

¹Docente de la carrera de Medicina de la Universidad Regional Autónoma de los Andes (UNIANDES Ambato), Ecuador

²Docente de la carrera de Medicina de la Universidad Regional Autónoma de los Andes (UNIANDES Santo Domingo), Ecuador

³Docente de la carrera de Medicina de la Universidad Regional Autónoma de los Andes (UNIANDES Ambato), Ecuador

Email: ua.juanviteri@uniandes.edu.ec; ua.juliorm92@uniandes.edu.ec;  ua.freddyjumbo@uniandes.edu.ec;

Abstract

The third most common disease worldwide, colorectal cancer (CRC) is responsible for around 10% of annual cancer diagnoses. The success of personalized treatment hinges on the ability to recognize biomarkers linked with CRC longevity and forecast the prognosis of CRC patients. The goal of this research is to provide a novel approach to doing multi-attribute colorectal cancer analysis by using machine learning algorithms with multi-criteria decision-making (MCDM) methods and neutrosophic set (NS). The NS is used to overcome the uncertainty in the dataset. This paper used the neutrosophic AHP method to get the weights of features in the used dataset. Then the machine learning algorithms are used to give analysis and prediction of colorectal cancer. The decision tree (DT) and support vector machine (SVM) is used to analyze and predict colorectal cancer. The dataset has nine features like age, gender, dukes stage, location, and Disease-free survival. This paper shows the analysis of the dataset and the correlation among the features.

Keywords: Machine Learning; AHP; MCDM; Neutrosophic Set; Colorectal Cancer.