As per WHO, in most cases, early diagnosis and screening of cancer increase the chances for successful treatment by focusing on indicative of patients as rapidly as possible. The medical diagnosis process is playing a significant role in the automatic analysis of the image and evaluation of the skin cancer, but in some ways, it gets detrimental and pain also. Keeping the last point in mind, we proposed a method for the analysis of image to determine whether it is skin cancer or not. The machine learning characteristic is to be considered as prospect theory, decision models that exhibited new data, and they are capable of adapting different conditions independently. Various methods have been proposed to deal with skin cancer disease. As cancer exists in different stages, premature diagnosis is necessary for a patient to control cancer at early stages. The steps that are included in this study are collecting the skin cancer image, processing by some skin cancer parameters. The tools that we have used to detect data from different skin cancers images include principal component technology, which reduce the high-dimensional image into small dimension and after that applied to enhance clustering of the data, which is the most important method of unsupervised learning. This chapter draws a novel hybrid statistical feature extraction by using GLCM and PCA technology and statistical dimension reduction scheme with fuzzy C-mean novel clustering. The main gist of evolving this paper is to extract the people to the deleterious diagnostic test technique and easily identify the cancer patients through this entire process of SVM classification algorithm with suitable appropriate results, in some pandemic position. The results of the proposed scheme give better performance in contrast with the existing popular methods.