In this paper, we try to evidence the benefits of using advanced ML for credit risk modeling based on real use cases, also shedding some light on the potential pitfalls related to the use of ML models in credit risk with a specific focus on interpretability issues and their possible solutions. This article was prepared
In this paper, we try to evidence the benefits of using advanced ML for credit risk modeling based on real use cases, also shedding some light on the potential pitfalls related to the use of ML models in credit risk with a specific focus on interpretability issues and their possible solutions. This article was prepared in association with Accenture (Meriton Ibraimi & Norbert Sommer).