Machine Learning Methods for Multiclass Brain Tumor Classification in MRI Scans: A Comprehensive Analysis

The brain is one of the most vital organs, and yet, it has become the second highest reason of death around the globe. The reasons for which being misdiagnosis, lack of physicians with the required expertise, and delay in decision-making. These problems can be easily resolved using machine learning approaches. In this paper, the performance of various machine learning algorithms is compared to identify which model can give the best performance to perform a multiclass classification of brain MRI scans. For this study, we used a dataset containing three types of tumors present in the brain. The performance of six machine learning algorithms is compared: SVM, decision tree, random forest, Naive Bayes, logistic regression, and neural networks. The performance of each of these metrics was later evaluated on the accuracy metric.