A variety of issues, including classification, link prediction, and graph clustering, have been solved using graph neural network (GNN), an efficient method for handling non-Euclidean structural data. Another effective and reliable mathematical tool for classification and regression applications is support vector machine (SVM). We hope that this paper will help readers gain a better knowledge of the latest developments in graph neural networks and how they are used in a variety of fields. We also describe current research on using support vector machines for prediction and classification problems. Following that, a comparison between SVM and GNN is made, and the results are discussed.