Matthias Jakobs

Matthias Jakobs focuses his research on trustworthy machine learning. He is currently working on various problems, including providing guarantees for explanation methods based on Shapley values, both in theory and in practical applications. Additionally, he is exploring the combination of explainability models with Bayesian Neural Networks (BNNs).

He is particularly interested in illuminating the decision-making process of black-box models to instill greater trust in the decisions made by neural networks, for both users and experts.