- Our work "Efficient Robustness Certificates for Discrete Data: Sparsity-Aware Randomized Smoothing for Graphs, Images and More" on certifiable robustness for ML models handling discrete data (e.g. Graph Neural Networks) has been accepted at the International Conference on Machine Learning (ICML), 2020. Our technique also handles sparse data, subsubmes other approaches, and at the same time has a significantly lower computational complexity (in particular, the complexity does no longer depend on the dimensionality of the input data; a major bottleneck of existing robustness techniques for discrete/binary data).
- Our work on scaling-up density based clustering "Gauss Shift: Density Attractor Clustering Faster than Mean Shift" has been accepted the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML-PKDD), 2020.
Congratulations to all co-authors!