A comparison of machine learning/deep learning algorithms for the classification and prediction of cancerous tumors

With cancers being one of the leading causes of death, research in this field has considerably increased over the past decades. One way to improve the recovery chances in patients is its early detection. Moreover, with the improvements of Machine Learning algorithms, the determination of early diagnosis is made easier. In this study, we compare the performance of different algorithms from Linear regression and classification tree to pre-trained models such as ResNet50 or VGG16. In addition, we develop our own Convolutional Neural Network. We show that in addition to a similar performance in terms of accuracy, our CNN model provides much faster prediction than the pre-trained models.