Panagiotis Misiakos

e-mail:
Office: CAB H 81.2
Universitätstrasse 6
Zürich, Switzerland

I am a PhD student of Markus Püschel. My research interests include applications of mathematics in Signal Processing and Machine Learning. Currently, I am working on DAG learning methods from a causal Fourier analysis perspective.

Publications

Submitted

Unrolled Möbius Total Variation for Denoising Sparse-Input DAG Signals
Panagiotis Misiakos and Markus Püschel
Submitted to Proc. International Conference on Acoustics, Speech, and Signal Processing (ICASSP) 2027.

2026

Causal Discovery in Nonlinear Time Series with Sparse Structural Shocks
Panagiotis Misiakos and Markus Püschel
To appear in Proc. Asilomar Conference on Signals, Systems, and Computers 2026.

Cyclic digraph with spectral radius constraint

Learning Stable Digraphs from Sparse-Input Linear Structural Causal Models
Panagiotis Misiakos and Markus Püschel
In The 42nd Conference on Uncertainty in Artificial Intelligence (UAI) 2026, Proc. Machine Learning Research 337.
[pdf] [PMLR] [poster] [slides]

Learning Directed Acyclic Graphs from Max-times Structural Equation Models with Sparse Input
Panagiotis Misiakos and Markus Püschel
In Proc. International Conference on Acoustics, Speech, and Signal Processing (ICASSP) 2026, pp. 6201–6205.
[IEEE Xplore] [poster]

2025

SpinSVAR: Estimating Structural Vector Autoregression Assuming Sparse Input
Panagiotis Misiakos and Markus Püschel
In The 41st Conference on Uncertainty in Artificial Intelligence (UAI) 2025.
[pdf] [poster]

Learning Time-Varying Graphs from Data with Few Causes
Panagiotis Misiakos and Markus Püschel
In Proc. International Conference on Acoustics, Speech, and Signal Processing (ICASSP) 2025.
[IEEE Xplore] [poster] [slides]

The CausalBench challenge: A machine learning contest for gene network inference from single-cell perturbation data
Mathieu Chevalley, Jacob Sackett-Sanders, Yusuf H Roohani, Pascal Notin, Artemy Bakulin, Dariusz Brzezinski, Kaiwen Deng, Yuanfang Guan, Justin Hong, Michael Ibrahim, Wojciech Kotlowski, Marcin Kowiel, Panagiotis Misiakos, Achille Nazaret, Markus Püschel, Chris Wendler, Arash Mehrjou, Patrick Schwab
Conference on Causal Learning and Reasoning (CLeaR), Proc. Machine Learning Research 275, pp. 1–19, 2025.
[arXiv]

2024

Learning signals and graphs from time-series graph data with few causes
Panagiotis Misiakos, Vedran Mihal, Markus Püschel
Oral presentation in Proc. International Conference on Acoustics, Speech, and Signal Processing (ICASSP) 2024.
[IEEE Xplore] [presentation]

2023

Learning DAGs from Data with Few Root Causes
Panagiotis Misiakos, Chris Wendler, Markus Püschel
Advances in Neural Information Processing Systems 2023.
[pdf] [poster] [short slides]

Learning Gene Regulatory Networks under Few Root Causes assumption
Panagiotis Misiakos, Chris Wendler, Markus Püschel
3rd prize award in GSK.ai CausalBench Challenge 2023, hosted in MLDD workshop ICLR 2023.
[OpenReview] [arXiv] [slides]

2022

Neural Network Approximation based on Hausdorff distance of Tropical Zonotopes
Panagiotis Misiakos, Georgios Smyrnis, Georgios Retsinas, Petros Maragos
In International Conference on Learning Representations (ICLR) 2022.
[pdf] [poster] [slides]

2020

Diagonalizable Shift and Filters for Directed Graphs Based on the Jordan-Chevalley Decomposition
Panagiotis Misiakos, Chris Wendler, Markus Püschel
Proc. International Conference on Acoustics, Speech, and Signal Processing (ICASSP) 2020.
[IEEE Xplore] [poster]

Talks

2026

Causality: From Combinatorial to Continuous Optimization
Computer Vision, Speech Communication and Signal Processing group, National Technical University of Athens. 2026

Learning Stable Digraphs from Sparse-Input Linear Structural Causal Models
The 42nd Conference on Uncertainty in Artificial Intelligence (UAI). 2026

2025

SpinSVAR: Estimating Structural Vector Autoregressions Assuming Sparse Input
Seminar for statistics group, ETH Zurich. May 2025

Learning Graphs from Structural Vector Autoregressions with Sparse Input
GSP workshop, Mila-Quebec AI institute. May 2025

2024

Learning Directed Acyclic Graphs from Data with Few Root Causes
Antonio Ortega's group, University of Southern California. Dec 2024

Supervised Students

Piotr Jasinski, Bachelor’s thesis, in progress
Where to Measure a DAG? Sampling and Recovery of Signals with Sparse Structural Shocks
Ayse Sen, Bachelor’s thesis, in progress
Learning DAGs by Searching over Topological Orders under Sparse Structural Shocks
Johannes Göttle, Bachelor’s thesis 2025
Learning linear SEMs with Transitive Closure on a Semiring
Dillon Martinelli, Bachelor’s thesis 2025
Learning Hypergraphs from Fourier-sparse Signals
Isabel Haas, Bachelor’s thesis 2023
GSP Graph Learning approaches applied to DAG Learning
Evangelos Pipis, Summer fellow 2023
Learning Directed Graphs with Cycles and Few Root Causes
Davide Bizzaro, Summer fellow 2022
DAG learning with SEMs on tropical semirings

Teaching

Fall 2026, Head TA in Information Systems for Engineers
Fall 2025, Head TA in Information Systems for Engineers
Spring 2025, Head TA in Information Retrieval
Fall 2024, Head TA in Information Systems for Engineers
Spring 2024, Head TA in Information Retrieval
Fall 2023, teaching assistant in Information Systems for Engineers
Spring 2023, teaching assistant in Information Retrieval
Fall 2022, teaching assistant in Algorithms and Data Structures
Spring 2022, teaching assistant in Introduction to Machine Learning

Education

Master (M.Eng.) in Engineering.
School of Electrical and Computer Engineering, National Technical University of Athens, Greece.
Thesis (in Greek) supervised by Prof. Petros Maragos.

November 2021