In this paper we address the problem of differentiating between malignant and benign tumors based on their appearance in the CC and MLO mammography views. Classification of clustered breast microcalcifications into benign and malignant categories is an extremely challenging task for computerized algorithms and expert radiologists alike. We describe a deep-learning classification method that is based on two view-level decisions, implemented by two neural networks, followed by a single-neuron layer that combines the viewlevel decisions into a global decision that mimics the biopsy results. Our method is evaluated on a large multi-view dataset extracted from the standardized digital database for screening mammography (DDSM). Experimental results show that our network structure significantly improves on previously suggested methods.