Human-Centric AI Webinar #3: Advances in Explainable AI

This webinar includes two research talks on recent advances in Explainable AI for complex black-box models. The first talk explores how feature-based attribution methods can help make deep neural network predictions more understandable, from scalar outputs and multimodal medical data to structured predictions, survival analysis, and predictive uncertainty. The second talk presents work on explainable cancer segmentation through classification, highlighting how XAI can support more transparent and interpretable medical AI systems. Together, the talks show how explainability can move AI beyond accurate predictions toward models that are more trustworthy, actionable, and useful in real-world decision-making.

Participation is free!

Date: 17 June 2026, 13:00-14:00 CEST
Location: Online (Teams)
To register, you must fill in the following form.

This webinar is part of a series of online AI events organised by ENFIELD, a project co-funded by the European Union.

Target audience

AI researchers, practitioners, and students with an interest in human-centric AI, explainable AI, or smart environments. Also relevant for professionals working on wearable health technology, multi-agent systems, or AI-assisted wellbeing and stress monitoring.

Speakers

Andrei Dugăeșescu, University POLITEHNICA of Bucharest
Niklas Koenen, BIPS – Leibniz Institute for Prevention Research and Epidemiology

Moderator

Chao Zhang, Assistant Professor, Eindhoven University of Technology, the Netherlands

Topics & Abstracts

Title:	From Deep Neural Network Predictions Toward Understanding: Advances in Explainable AI for Complex Black-Box Models
Speaker:	Niklas Koenen
Abstract:	Machine learning models, particularly deep neural networks (DNNs), achieve impressive predictive performance on high-dimensional and multimodal data, yet their decision-making remains hidden inside the “black box.” Explainable AI (XAI) addresses this challenge through feature-based methods that reveal which inputs are decisive for a model’s output. This talk presents recent advances in two parts, organized around the input and output sides of the attribution problem. The first part concerns the input side: attributing scalar predictions back to features. While many attribution methods already exist, especially for DNNs, the work here is less about adding new ones and more about understanding them and why they often disagree, making them broadly accessible to applied users, and applying them to multimodal medical data, e.g., for the early detection of cognitive impairment. Going further, a conditional variant of feature importance based on generative models considers the dependencies between input features and thereby captures each feature’s unique contribution given the others, rather than its marginal effect. The second part turns to the output side, where the prediction itself has structure beyond a single scalar. Attribution is first extended to survival analysis, where the prediction takes the form of a survival function over time. While this treats the multivariate outcome pointwise, the most recent work extends the output-side perspective from time-varying predictions to the full predictive distribution and explains the model’s predictive uncertainty itself. A hierarchical entropy-based Shapley framework attributes the uncertainty of multi-step probabilistic forecasts to individual features, decomposed into marginal, sequentially conditional, and joint contributions across output components, plus a cross-component term that captures how features induce temporal dependence in the output. Taken together, this talk works toward opening the black box from both sides by explaining what modern ML models predict and how confident they are about it, while accounting for dependencies on both the input and output sides.