L. Bermúdez Morata, D. Anaya Luque, J. Belles-Sampera
Explainable Artificial Intelligence (xAI) plays a crucial role in enhancing our understanding of decision-making processes within black-box Machine Learning (ML) models. Our objective is to explain various xAI methodologies, providing risk managers with accessible approaches to model interpretation. To exemplify this, we present a case study focused on mitigating surrender risk in insurance savings products. Initially, we utilize real data from universal life policies to build a logistic regression and tree-based models. By employing a range of xAI techniques, including a novel Kohonen Neural Network (KNN) of Shapley values, we gain valuable insights into the inner workings of the tree-based model. Finally, we focus on understanding the factors driving the model’s performance and the analysis of the different risk profiles present in the portfolio in terms of surrender risk mitigation.
Palabras clave: Machine Learning, Shapley values, Kohonen Neural Network
Programado
AR1 Análisis de Riesgos I
12 de junio de 2025 11:30
MR 1