Tamás Varga (LinkedIn)
In the thesis, the author attempted to predict the severity of acute pancreatitis using a graph-based neural network that employs an attention mechanism in order to present a framework and deep learning model that enables the predictive analysis of tabular data using a graph-based approach. The thesis describes in detail the methodology and application of the Graph Attention Network model, with a particular focus on presenting the most optimal parameter settings and decision points for the research problem. In order to put the results of the neural network-based analysis into context, the author analyzed the research problem using machine learning methods and compared the results with those of the deep learning algorithm. As a result, the thesis contributes to the methodological and practical understanding of the deep learning model based on the attention mechanism discussed.