One of the most common cancers, skin cancer accounts for roughly half of the cancer prevalence in worldwide. Melanocytes, which are cells that produce pigment, are the source of melanoma, a kind of skin cancer. Both the incidence and mortality rates of melanoma have sharply increased in recent years. The ailment that attracted a variety of studies can be treated more effectively by medical professionals with early identification. This article focuses on several segmentation techniques and a classification strategy based on machine learning for the identification of melanoma. The melanoma areas are easily recognized by the segmentation methodologies, and the classification procedure is made simpler. Convolutional neural networks (CNN) with U-Net are used for segmentation. A support vector machine (SVM) with a variety of kernels and a k-nearest neighbor (KNN) with a variety of k values are used to classify the segmented image in the classification phase. Performance demonstrates that the segmentation process significantly increases the SVM's classification effectiveness when acquired values from the machine learning classifier are compared. Further the results shall be improved by enhancing the analysis with large dataset.