The defining feature of cancer is the unregulated growth of atypical cells, which can potentially spread to other organs or tissues. One of the most dangerous and fatal types of cancer is skin cancer. Fortunately, early detection allows for effective treatment of skin cancer. The effectiveness of treatment and the survival rates of patients hinge on timely and accurate diagnosis. It is very uncommon for dermatologists’ subjective opinions to color their evaluations of skin lesions, which may lead to inaccurate diagnoses or unneeded treatments. Recent developments in machine learning and deep learning have empowered medical professionals to detect skin cancer earlier and with greater accuracy, ultimately reducing costs for patients and preventing expensive diagnostic procedures. The diagnosis and classification of skin cancer mostly make use of convolutional neural networks (CNNs) and other deep learning algorithms, with some contributions from hybrid approaches and machine learning techniques. Being very powerful classifiers, these methods show great promise for the early diagnosis of skin lesions. Skin cancer detection methods were the focus of this investigation. Techniques from Machine Learning and Deep Learning are included in the strategies. Our examination of skin cancer detection techniques also encompassed an analysis of transfer learning and hybrid methodologies.