Melanoma is a highly malignant form of skin cancer, appearing in minor modifications of size, shape, and color; hence making early detection and correct assessments tough. Melanoma arises from melanocytes, which synthesize skin pigments, and is mainly induced by UV light exposure, with more severe effects on individuals with low melanin content. Melanoma is documented to have 325,000 new cases and 57,000 deaths estimated globally in 2023. Conventional diagnosis mainly depends on visual inspection and is subject to human error. In contrast, machine learning comprises a revolutionary approach by rendering melanoma detection and measurement automatic and thus fast and accurate. This study concentrates on regression models using machine learning, like Neural Networks, Random Forests, Support Vector Machine (SVM), Linear Regression, and Ensemble Methods, for predicting tumor size. The results demonstrated the predictive ability and validity of machine learning models in the evaluation of melanoma. The study highlights the application of AI in dermatology concerning its potential to assist clinicians in making quantifiable, data-driven diagnosis decisions leading to early detection and optimal treatment planning.