One of the most common types of cancer found among women is breast cancer and it is associated with a high mortality rate. Medical imaging remains a dependable method for detecting breast cancer, but manual image interpretation is time-consuming. This research work introduces a novel deep learning approach utilizing Convolution Neural Networks (CNN). CNNs are commonly employed for image classification, although identifying precise hyperparameters and architectures presents a significant challenge. The work included developing a highly accurate CNN model specifically designed for the detection of breast cancer using mammography. The method proposed relies on the evolutionary algorithms to search for appropriate hyperparameters and CNN model for classification. The experimental outcomes demonstrated that the suggested CNN model delivered superior accuracy compared to another research conducted in the field on tuning the hyperparameters by evolutionary algorithms, specifically particle swarm optimization, whale optimization, and gray wolf optimization. The proposed approach can be regarded as a potent method for predicting breast cancer.