Detecting brain tumors from MRI images is a critical task in medical imaging and plays a crucial role in early diagnosis, treatment planning, and patient care. Image processing techniques rely on handcrafted features, which might not capture the full complexity of brain tumors. They may struggle to represent subtle and discriminative features necessary for accurate detection. The proposed model focuses on the feature extraction using the ensemble ranking approaches. The model proposes “Kernel methods”, such as Multiple Kernel Learning (MKL), can be employed to combine features from multiple kernels or similarity measures, each obtained from different feature extraction methods. The extracted features are reduced further using the CNN and classified using the Random Forest algorithm. The main advantage of MKL is it can automatically learn the relevance of each kernel and perform feature selection by assigning appropriate weights to them. The learning process allows the model to focus on the most informative features while discarding less relevant ones. The integration of MKL features with CNN takes advantage of the diverse and informative representations provided by MKL and the hierarchical feature learning capability of CNNs. This approach can be particularly beneficial when dealing with multi-modal data, data from different sources, or complex tasks where the combination of complementary features can lead to improved performance and generalization.