Detection of Skin Cancer using Machine Learning

The paper presents the design and development of concept for detecting the skin cancer using AI and ML approaches. Over the last few years, Deep Learning has been on the spotlight when it comes to solving many complex problems, namely Convolutional Neural Networks (CNN) in the field of image recognition. Since this Neural Networks can sometimes achieve more than 50 layers deep, Transfer Learning (TL) techniques are also getting popular to retrain some of the layers of Neural Networks that learned from a generic dataset, putting it to use in another case. In this work, we attempt using CNNs and TL techniques on pre-trained Neural Networks to distinguish melanoma skin cancer from seborrheic keratosis and nevus benign tumours. The work presented in this paper is the mini-project work of the undergraduate students of the engineering college guided by their supervisors.