Prediction and classification of skin melanoma cancer using active hybrid machine learning technique

Abstract

Melanoma is a dangerous skin cancer that may spread if not detected early. Thus, the medical industry has evolved with automated diagnostic tools that may assist physicians and even regular people diagnose an illness. Here, we provide a combined method for identifying melanomas on the skin. For our proposed approach, we make use of three forecasts. This problem was solved by training a neural network and two-learning machine classifiers on data describing the borders, textures, and colours of skin lesions. These tactics are combined for maximum effectiveness by majority voting. Tests show that using all three methods simultaneously improves precision.