“…Therefore, the BreakHis dataset [2, 4, 9, 12, 14, 20, 23, 27-29, 31, 32, 40, 47, 48] and WBCD dataset [1,5,10,17,21,21,34,41,42] can be the most suitable dataset with a sufficient number of breast cancer images. SELF ❼ The researchers have adopted machine learning [1,6,10,24,29] or deep learning [2,4,12,14,18,28,31,40,47,48] or ensemble learning [3, 5, 17, 20, 21, 23, 27, 30, 32-34, 36, 41, 42] techniques to address the breast cancer classification problem, and put their best efforts to improve the performance of their proposed approach(s). From Table 1, we observe that most of the existing works have mainly borrowed the deep learning techniques to address the breast cancer classification problem because of image datasets [2,4,9,12,14,18,20,23,28,31,32,40,47,47,48].…”