A hybrid of CNN and PCA in skin cancer for improving accuracy detection

Cancer is a generic term for many diseases that can affect any part of the human body. Skin cancer is considered to be the greatest common and dangerous type of cancer. Information technology techniques are required to detect and diagnose skin cancer. Therefore, there is a essential for an early and accurate diagnosis and treatment of skin cancer using effective and deep learning techniques. This research work proposes automatic diagnosis of skin cancer by employing Deep Convolution Neural Network (DCNN). The distinguishing feature of this research is it employs DCNN with 12 nested processing layers increasing the diagnosis and detection of skin cancer accuracy. Moreover, it can detect the cancer if it is malignant or benign in the early stages with high accuracy. Beside neural network, machine learning techniques of Naive Bayes and random forest are also utilized to detect skin cancer. This research work results concluded that the deep learning technique are more effective than machine learning in terms of skin cancer detection. By applying Naive Bayesian on the proposed system accuracy of 96% were achieved, similarly for Random forest method, an accuracy of 97% were achieved. The accuracy of 99.5% were achieved by applying Deep CNN network.