INTERPRETABILITY OF MACHINE LEARNING MODELS ORIENTED TO BINARY CLASSIFICATION PROBLEM
In this project, different model-agnostic interpretability strategies were implemented and analyzed in the context of the binary classification problem, in order to develop a more thorough understanding of the Machine Learning (ML) models involved. As such, local and global methodologies were addressed to comprehend variables dynamics behind a complete black-box model.
Contract holder: Santander España
Performance Term: April 2020 – July 2020