The classification of brain tumors in magnetic resonance imaging (MRI) is very significant for diagnosis and detecting the tumors in the medical world. The major advantages of MRI are soft tissue analysis and non-invasive but the drawback is a long time consumed by a medical expert to draw conclusions. The classification algorithms help in order to overcome the disadvantage of MRI. In this work, brain tumor classification is done based on the GLCM feature values where 165 brain medical images are presented. The main aim of this work is solving the cancer classification by implementing 3 different classifiers such as NB (Naïve Bayes), NN (Neural networks), and SVM RBF (Support vector machine Radial Basis Functions). The features of the images are extracted using GLCM and further, the feature values are classified as either affected or not affected. Among 3 classification approaches, NN achieves superior accuracy (99%) with less error rate.