Paper Title
Effects of Data Augmentation in Melanoma Cancer Classification using Convolution Neural Network

Abstract
In deep learning, sufficient training data is required for the training of the image classification model. The increase in the number of data is an important factor that increases the success of the model, since it will allow the model to recognize the different data especially during the test phase. However, it may sometimes be difficult to obtain sufficient data for the training of the model in the real world. Therefore, data augmentation is a very important method in increasing the success of classification and test using the existing data set as reference data. In this study, five different data augmentation methods were used: Adding Noise, Cropping, Flipping, Rotation, and Translation. As a deep learning model, Alex Net structure was preferred for classifying data. The data sets obtained by these methods were compared with the original data set under equal conditions. It was clear that the results of all data enhancement methods are more successful than the results obtained using the original data set. These results clearly show that data augmentation techniques increase both training and test successes. Keywords - Convolution Neural Network, Data Augmentation, Image Classification, Deep Learning