Dimosthenis Pasadakis

I am a faculty member (lecturer of mathematical engineering) at the University of Applied Sciences and Arts of Southern Switzerland (SUPSI) in the Department of Innovative Technologies. I am also a researcher at Università della Svizzera italiana (USI) in the Institute of Computing. The focus of my research is centered around HPC algorithms for graph learning and combinatorial and continuous optimization for graph clustering and anomaly detection.

In the private sector, I act as the Chief Operating Officer (COO) of Panua Technologies. We develop customized high-end software solutions for large-scale prediction, simulation, optimization, and graph analytics.

My lecturing, research, and industrial activities are based in Lugano, in the south of Switzerland.

logos_full

For students

Emerging topics in advanced computing, Phd course, USI.
Data analytics for fraud detection, MSC, MFT, PhD course, USI.
Numerical analysis, BSc in Data Science and AI, SUPSI.
Calculus II, BSc in Data Science and AI, SUPSI.
Probability and statistics, BSc in Data Science and AI, SUPSI.
For MSc and BSc thesis topics in SUPSI and USI, please contact via e-mail.

Adjacency matrix estimation via maximum likelihood

❮ ❯

Nonlinear PageRank for local graph clusters

❮ ❯

Directed spectral methods for topological ordering

❮ ❯

09/26: I am joining SUPSI as a faculty member (lecturer of mathematical engineering).
06/26: I am visiting the Complexity Science Hub in Vienna, the first week of June for an invited talk.
05/26: Congatulations to Claudio for winning the Best Presentation Award at Premio Swissengineering Ticino 2026, for his MSc thesis Using time-series clustering for a better financial crime segmentation model.
04/26: I presented An introduction to linear and non-linear spectral clustering at Tandy Warnow's class Graph Algorithms for Community Structure Detection in Large Networks at the Siebel School of Computing and Data Science, University of Illinois Urbana-Champaign.
03/26: We organized the minisymposium Scalable numerical algorithms and graph analytics for large-scale applications at the SIAM PP26 in Berlin.
02/26: Invited talk at the Aristotle University Research Dissemination Center.
12/25: Presenting topics associated to supervised and unsupervised learning connected with the Governance and development of AI, in the Academy & Finance workshop in Geneva and Zurich.
11/25: Presenting our preprint on spectral topological orders, in the workshop Networks for Science in Berkeley, California.
09/25: Congratulations to Xiaohe and Georg for bringing home the Best poster award, associated with the IEEE Cluster 2025 paper.
06/25: I will be presenting our work this summer at VCMF 25 (Vienna), SIAM FM 25 (Miami), and at the Duke Computer Science Colloquium (Durham).
03/25: I will be visiting the Huawei Research Center Zürich in March.
03/25: The second edition of the AI & compliance seminar is taking place in Zürich.
11/24: I am presenting the foundation of the AI tools used in compliance at the Academy & Finance seminar in Geneva.
10/24: I will be presenting our recent work at the Deloitte Zürich seminar The power of graph technology in Financial Crime.
09/24: Our article Multilevel diffusion based spectral graph clustering has received the Outstanding Paper Award at IEEE HPEC.
06/24: Our interview with Albert-Jan Yzelman at ACM PASC regarding high performance graph analytics is now available online.
06/24: The poster accompanying our article GAMLNet: a graph based framework for the detection of money laundering has received the Best Poster Award at IEEE SDS24.
03/24: Our article Sparse Precision Matrix Estimation With SQUIC is published in ACM Transactions on Mathematical Software.
03/24: We are organizing the minisymposia "Learning and Clustering Tasks on Graphical Structures" at SIAM LA 24 in Paris, and "High Performance Graph Analytics" at PASC 24 in Zürich.
11/23: I have been awarded a DDSA grant to visit the Department of Mathematical Science of Aalborg University.
09/23: Our article Nonlinear spectral clustering with C++ GraphBLAS has received the Outstanding Short Paper Award at IEEE HPEC.
04/23: Our article Sparse Quadratic Approximation for Graph Learning is published in IEEE Transactions on Pattern Analysis and Machine Intelligence.
04/23: The Swiss National Science Foundation (SNSF) project Balanced Graph Partition Refinement Using the Graph p-Laplacian that supported my Phd studies is now complete.
03/23: I have successfully defended my Phd thesis entitled Learning and clustering graphs from high dimensional data.
07/22: Chairing the AP1B - ACM Papers Session 1B in PASC'22.
12/21: Our article Multiway p-spectral graph cuts on Grassmann manifolds was featured in the newsletter of the National Centre for High Performance Computing of the University of Erlangen (NHR@FAU).