 The proposed methodology is able to achieve higher accuracy and efficiency compared to other methods in skin cancer detection. It uses a MobileNet V3 architecture to extract relevant image representations, which are then fed into a modified version of the Hunger Games search algorithm. This algorithm selects the most relevant features to maximize the model's performance. The proposed method was evaluated on the ISAC-2016 and PH-2 datasets, with accuracy rates of 88.19% and 96.43%, respectively. This article was authored by Abdel Ghani Dahu, Ahmad Owassiri, Al-Hassan Mabruk, and others.