(2022-07-14) Georgiadis, Theofilos; Γεωργιάδης, Θεόφιλος; Athens University of Economics and Business, Department of Management Science and Technology; Lekakos, Georgios; Mourtos, Yiannis; Korfiatis, Nikolaos
In this thesis we will test many different machine learning models for credit scoring. Credit scoring is the technique used to predict the probability of default for a client that wants to take a loan and decide if he/she should be granted the loan or not. Firstly, we will test the performance of the different models compared to the traditional model used in credit scoring which is the logistic regression to see if we can achieve better performance using more advanced models. Then we will explain the predictions of the models to see how each feature contributed to the prediction, which is a very important aspect in this sector, which is highly regulated.
