Model Explainability

This folder contains examples and best practices, written in Jupyter notebooks, for explaining and interpreting models. Being able to explain and understand machine learning models not only helps guiding further model improvements, but more importantly, it’s critical for gaining users’ trust in the models and detecting biases caused by the training data.

Summary

Notebook	Environment	Description	Dataset	Language
DUUDNM	Local	Interpreting DNN Layers using Mutual Information.		en