The manual diagnosis of tumor is a time intensive process which may result in errors on the part of humans, leading to erroneous detection and classification of tumor types. The contributors provide a structure based on machine learning for recognizing different kinds of brain tumors, aimed at streamlining the diagnostic process for healthcare providers by integrating complex procedures related to medicine. The characteristics of the segmented tumor region are obtained, and the CNN classifier approach subsequently employs the tumor features to determine the tumor existence. In order to put the suggested model into action, pre-trained a Deep Convolutional Neural Networks (DCNN) are utilized. These designs are each coupled with a unique classifier. Using the architectures as transfer learning methodologies, the characteristics from the pre-trained DCNN framework are extracted, and then the Support Vector Machine (SVM) classifier is used to classify them. This process is carried out by deploying the models. The accuracy of the system that detects tumors automatically will be elevated as a result of utilizing this strategy. In the context of Magnetic Resonance Imaging (MRI), methods for augmenting data are utilized in order to stop the network from being overly fitted. When splitting MRI images, using several networks is examined by contrasting the outcomes with those obtained using just one network. The severity of the brain tumor is accessed by using the Convolutional Neural Network method, which yields trustworthy results.