A Brain Tumor (BT) is the leading cause of mortality in the world, and it is the most prevalent type of tumor, affecting persons of all ages. However, it is a curable kind of cancer if discovered early. The biopsy is used to categorize BTs, but it is rarely performed before the removal of the tumor by brain surgery. For speeding up the treatment procedure, minimizing unnecessary surgery, and reducing the likelihood of diagnostic errors caused by human intervention, an image-processing approach for tumor is crucial. Machine learning (ML) and other technological advancements are helping radiologists to diagnose tumor from Magnetic Resonance Imaging (MRI) scans by a non-invasive method. To distinguish between three kinds of BTs and healthy brain tissue in MRI, this research introduces a hybrid architecture. This research proposes a hybrid ML classification algorithm based on Support Vector Machine (SVM) and Gaussian Nave Bayes (GNB). After being trained, the suggested method was put to the test using a collection of magnetic resonance imaging (MRI) data consisting of 100 images of glioma, 115 images of meningioma, 74 images of pituitary tumor, and 105 images of normal brain tissue. The results of the hybrid model are compared with the standalone ML models like SVM, GNB, Linear Discriminant Analysis (LDA), and Random Forest (RF). According to the findings, the hybrid model outperforms the other ML models in terms of accurate classification while maintaining a low false rate.

Brain Tumor, Hybrid Model, Medical Images, Feature Extraction, Augmentation, Machine Learning.