Early detection of breast cancer plays a vital role in stopping the spread of disease and reducing mortality rates among women. Radiologists face the arduous task of manually interpreting mammograms, which can lead to potential diagnostic errors. Therefore, it is imperative to employ a Computer-Aided Diagnosis (CAD) approach to automatically identify early-stage breast cancer and aid radiologists in their decision-making process. In this paper, we propose a deep learning based automatic breast cancer classification method which utilizes the hybrid features extracted from two different methods Gray-Level Co-occurrence Matrix (GLCM) and Convolutional Neural Networks (CNN). We merge the features before feeding them to the model for classification. Eventually, we evaluate the model performance using the dataset of breast mammography images with masses and compare its performance with state-of-the-art.