It is one of the challenging tasks in medical imaging where machine learning algorithms have been increasingly used in diagnosis: the identification of brain tumours. The paper compares the interpretability and performance of the logistic regression classifier and support vector machines classifiers in the identification of brain tumours. We use a dataset of images obtained through MRI, including examples of those diagnosed with a brain tumour and others who are not. Using the SVM and LR models, the data is then preprocessed to elicit pertinent features using a range of kernel functions and regularization approaches. Performance of the models in making such predictions is evaluated through use of various performance measures such as area under the receiver operating characteristic curve, sensitivity, specificity, and accuracy. Our results indicate that both SVM and LR models have good performance in the identification of brain tumours but with slight superiority of SVM over LR models regarding accuracy and AUC-ROC, respectively. On the other hand, interpretability is better with LR models, since they explicitly illuminate the importance of some elements in the classification process.