期刊名称：Documents de Travail du Centre d'Economie de la Sorbonne
印刷版ISSN：1955-611X
出版年度：2018
出版社：Centre d'Economie de la Sorbonne
摘要：Due to the hyper technology associated to Big Data, data availability andcomputing power, most banks or lending nancial institutions are renewingtheir business models. Credit risk predictions, monitoring, model reliabilityand e ective loan processing are key to decision making and transparency.In this work, we build binary classi ers based on machine and deep learningmodels on real data in predicting loan default probability. The top 10 impor-tant features from these models are selected and then used in the modellingprocess to test the stability of binary classi ers by comparing performanceon separate data. We observe that tree-based models are more stable thanmodels based on multilayer arti cial neural networks. This opens severalquestions relative to the intensive used of deep learning systems in the en-terprises.
关键词：Credit risk; Financial regulation; Data Science; Bigdata; Deep;learning
