Machine Learning with Multi-Criteria Decision Making Model for Thyroid Disease Prediction and Analysis

Abstract

This study adopted a decision system model that includes machine learning (ML) and multi-criteria decision-making (MCDM) for thyroid prediction and analysis. Many people face thyroid disease, so the early prediction of this disease can aid people around the world in early treatment. This paper integrates the ML algorithms with the MCDM methodology. Three ML algorithms are used in this paper: logistic regression (LR), support vector machine (SVM), and random forest classifier (RF). These algorithms are used to predict and analyse thyroid disease. The results show the RF has the highest accuracy, precision, and F1 score. The RF has 0.95 accuracy. The SVM has a 1.0 recall score. Then, the MCDM methodology is used with various criteria to rank and use the best ML algorithm. The TOPSIS method is used as an MCDM method to rank the ML algorithms. The mean method is used to compute the criteria weights. The results of the MCDM methodology show that RF is the best ML algorithm in this paper, followed by SVM, and the worst ML algorithm is LR.