We are happy to announce that our work From Link Prediction to Forecasting: Addressing Challenges in Batch-based Temporal Graph Learning has now been published in the journal Transactions on Machine Learning Research (TMLR). In our work, we highlight fallacies in the common batch-based evaluation of link prediction methods for temporal graphs.
This week, I will present three of our recent works at NetSci-X in Auckland, New Zealand. A first talk introduces our work on integrating statistical ensembles with GNNs. A second talk presents our recent work on the expressivity of temporal GNNs. In a third talk, I will present our work on activity cascades in OSS communities.
Next week, I will give a school lecture introducing Deep Graph Learning at NetSci-X in Auckland, New Zealand.
On Wednesday, I will give an invited talk at the Chair of Data Science in the Economic and Social Sciences at University of Mannheim. In my talk, I will present our recent work on temporal graph isomorphism and the expressivity of temporal GNNs.
I wish everyone a merry christmas and a good start into a healthy and successful new year 2026!
This week, we are presenting three works at this year's Learning on Graphs Conference Phoenix, Arizona, USA.
Next week, I will give a talk at Princeton University in New Jersey, USA.. In my talk I will present our recent work on the expressivity of causality-aware graph neural networks for temporal graphs.
Today, I will give a talk in the seminar of the Department of Computer Science and Information Engineering of the National Cheng Kung University in Tainan, Taiwan..