University of Twente Student Theses

Login
As of Friday, 8 August 2025, the current Student Theses repository is no longer available for thesis uploads. A new Student Theses repository will be available starting Friday, 15 August 2025.

Using a Graph neural network to predict loan defaults

Wasser, D.M. (2025) Using a Graph neural network to predict loan defaults.

[img] PDF
166kB
Abstract:A person not being able to pay their loan is a significant risk for banks. If this so-called ’loan default’ can be predicted, banks and other loan providers can better manage this risk. While a single loan default does not pose a problem for banks, multiple ones do. If loans are connected (for instance based on the location of the loan), and one of the connected loans defaults, the other ones could be more prone to defaults as well. So, in this paper, it is researched if a graph convolutional network (GCN) can be used to predict the probability that a loan will default. The connection between loans is mapped by converting a dataset to a graph and then inserting that graph into the GCN. The GCN is used to predict the probability that a loan will default. The used GCN is compared to two benchmark algorithms, namely a logistic regression model and a random forest model. The results of this paper show that the GCN performs better than the two benchmark algorithms, indicating that a graph structure offers a positive impact in predicting loan defaults.
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:54 computer science
Programme:Business & IT BSc (56066)
Link to this item:https://purl.utwente.nl/essays/107417
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page