Skin-Cancer-Detection

The potential for artificial intelligence to disrupt healthcare is incredible. I leveraged a specific type of machine learning algorithm, convolutional neural networks (CNNs), which is widely used in the area of image classification to detect skin cancer from pictures of moles. To decrease computational power needed to train the model, I used a pretrained convolutional neural network, Mobile Net which is trained on a dataset of 14 million images. The dataset used to train the model consist of 10,015 labelled images of skin moles seperated into 7 different clasess of skin cancer.

Content

• My article about CNNs