Explainable Machine Learning

Project duration:

Oct 2018 - Sep 2022

Explainable Machine Learning

Daily life is increasingly governed by decisions made by algorithms due to the growing availability of big data sets. Many machine learning algorithms, and neural networks specifically, are black-box models, i.e. they give no insight into how they reach their outcomes which prevents users from trusting the model. If we cannot understand the reasons for their decisions, how can we be sure that the decisions are correct? What if they are wrong, discriminating or amoral?
This project aims to create new machine learning methods that can explain their decision making process, in order for users to understand the reasons behind a prediction. Those explanations enable the user to check for correctness and robustness, and can also be useful for knowledge discovery.

Project Leader: