Comparative Analysis of Machine Learning Algorithms in Detection of Brain Tumor
Shahriar Hassan, Ali Ahnaf Hassan, Inan Marshad, M. Hosain, Musarrat Amin, Fahim Faisal, Mirza Muntasir Nishat
Abstract
In this paper, a machine learning approach is proposed to detect the presence of a brain tumor at the initial stages. The data extracted from the MRI scans of an affected person can be incorporated in various machine learning algorithms to facilitate the process. Ten algorithms were run here and the results obtained were extensively compared using the parameters: accuracy, precision, sensitivity, specificity, F1score and ROC-AUC. Among the proposed algorithms, Gradient Boosting, Random Forest and AdaBoost were found to be the most promising algorithms. But Gradient Boosting algorithm aced the rest with an accuracy of 98.78%, sensitivity of 99.3% and specificity of 95.2% while AdaBoost outperforms others in terms of precision and Random Forest shows the highest F1-score. Google Colab platform was used for running the algorithms.