Early Brain Tumor Analysis Through Hybrid Machine Learning Classification with Transfer Learning Approach with Hypothesis Function

Brain imaging through Magnetic Resonance Imaging (MRI) remains the foremost step in the detection and delineation of an atypical brain mass, which is a prerequisite for the examination of brain tumors. To tackle this a new Brain Tumor Hybrid Model with Transfer and Hybrid Machine Learning Techniques (BTHMTH) model is a brain-tumor MRI-based image classification system-worthy of banding as an invention. The N4ITK module uses a preprocessing stage to fix bias field distortions and make the image more uniform. The research is based on a specially created dataset that distinguishes between meningioma, glioma, and pituitary tumors. The features extracted from the DenseNet framework are converted into a hybrid learning model that has such classifiers as Support Vector Machine (SVM), XGBoost, and Gaussian Naive Bayes (GNB) to which ensemble methods are additionally appended. To tighten up the decision borders for tumor classification, the Minimum Message Length (MML) hypothesis function is employed. BTHMTH loss, accuracy, SVM confusion matrix, Gaussian NB, and XG Boost are some of the performance measures used by BTHMTH.