Detection of Malignancy On Dermis Using J48 and Random Forest Classifiers

The skin cancer has drawn tremendous medical interest over few years. Several skin abnormalities were not identified properly from statistics. The skin cancer detection is depended on texture variation of a specific location of skin. It is very difficult to recognize the malignant skin texture in the presence of common deformity viz., sunburns, eczema. So, in this paper, it has been proposed to detect melanoma from the dermoscopic patterns with skin cancer data set. The watershed segmentation technique is applied and segments are extracted that are subjected to features. The features extracted are classified using J48 and (RFC) Random forest classifiers and acquire best outcomes for skin lesions.