Interpreting magnetic hysteresis curves is a challenging and time-consuming task for researchers that can be automated using deep learning. While deep learning algorithms have led to breakthroughs in various scientific fields, only a handful of studies have attempted to integrate deep learning into the study of hysteresis curves, and more specifically, the classification of their plots. In this study, we investigated 584 DAT files, each describing four different magnetic hysteresis curves, through two separate preprocessing approaches. One approach required standardizing and interpolating the data. The other approach required normalizing, plotting, and saving the plots as grayscale images. Each process was evaluated through the training of a multilayer perceptron (MLP) and a convolutional neural network (CNN), respectively. The average accuracy and average loss obtained from training showed that the CNN performed significantly better than the MLP. The performance of the CNN indicated that the preferred approach for processing magnetic hysteresis data is through the generation of grayscale images.