Performance analysis of various deep learning techniques for brain tumor classification

Brain MRI image tumors are classified using machine learning techniques in which features are extracted and given to the classifier for the classification task. The manual extraction of elements is time-consuming and leads to poor performance due to a poor selection of features. This paper describes the performance analysis of various deep-learning techniques for brain tumor classification. These methods were assessed using three different categorization performance indices. The logistic regression and hybrid approach discovered a maximum classification accuracy of 89% for small and 87% for large datasets.