Breast cancer primarily affects women and is the second leading cause of mortality worldwide after lung cancer. Early detection of the mitotic stage at the cellular level can be a possible solution for timely curing it, as detection and counting of mitosis aid in determining how aggressive or malignant the cancer can spread. In the present research work, the author has proposed a deep neural network model for its performance visualization in classifying mitotic cells in histopathological images for breast cancer diagnosis. For this classification assessment, a dataset comprising 749 mitotic cells in the dataset's 1200 high-resolution histopathology images has been used. A transfer learning-based deep neural network model called squeezenet has been employed for this experimentation, which is a well-known pertained model that builds a solution for the current problem using the frozen weights stored in it while learning similar imagenet problems. In comparison to manual methods, deep learning findings produced better accuracy, recall, precision, and F-measure values of 86.84%, 85%, 93%, and 89%.