This video is hosted by QuantUniversity. Check out our upcoming courses for the QuantUniversity Winter School 2021 at: https://quantuniversity.com/courses.html
In this talk, Dr.Agus Sudjianto from Wells Fargo will unwrap the enigma of Deep Neural Networks and illustrate how the black box of deep ReLU networks can be simplified through exact local linear representation. He will share a real case study in credit risk assessment and also share Aletheia, a Python-based open-source package for interpretability. (https://github.com/SelfExplainML/Aletheia)
You can try examples on the Qu.Academy here: https://academy.qusandbox.com/#/market?tagId=5f06326c55fd05416e7fd28a
