Schizophrenia is a serious mental disorder which makes a patient abnormal than other patient thinks he is not there and everyone is for his enemy. As a result, it is so much important to detect the disease at an early stage. If we can detect the disease at an early stage, we can make the patient’s life normal. CNN (Convolutional Neural Network) -based technique for classification of the disease is used many times. In our research, we are using two class one is normal class, another is Schizophrenia class which is used transfer learning approach for classifying Schizophrenia disease from brain MRI data. In our presented method, our technique, which is based on transfer learning theory, uses a pre-trained MobileNet method to identify brain MRI images by extracting features using the sigmoid classifier method with a mean classification accuracy of 93.95%. Our proposed method exceeds all previous strategies. We utilize the Kaggle dataset to evaluate our technique. One of the important performance indicators used in this study is precision, recall, and F-score. Our classification method got accuracy of 90.62%.