Skin Cancer Diagnosis: Integrating CNN and Machine Learning Models

Abstract

Skin cancer is serious health concern that arises from abnormal growth of skin cells. It encompasses of three types, melanoma being the most aggressive form of cancer. The skin cancer starts to arise globally so accordingly there is being necessary public awareness, prevention for this, and strategies for early detection. Accurate diagnosis treatment and Early detection are better for effective management which will improve accuracy and efficiency for diagnostic process. Also, the Machine learning and deep learning algorithms we research and analyse for better outcomes and for innovation in fields of skin cancer detection. And for better treatment for patients, to reduce healthcare costs and overall management of skin cancer. Better constituency and through ongoing research we can try to overcome with help of future advancements in technology and the learning algorithm model. According to our reports and research we did through the dataset of skin images ML and deep learning we got 73% through Naïve Bayes 90% accuracy through random forest algorithm and 89% through CNN model algorithm. We do not conclude that Machine learning algorithm can be better since there are many other factors use in different algorithms so results are not based on high accuracy reports. Keywords: Deep learning, CNN model, Machine learning, Skin cancer