معرفی
Robert Peharz is an Assistant Professor at Graz University of Technology, where he leads research at the Institute of Machine Learning and Neural Computation. His work focuses on probabilistic machine learning, with particular emphasis on tractable probabilistic models, causality, and neurosymbolic AI.
Education and Career
- PhD from TU Graz (Austria) in 2015
- Postdoc at Medical University of Graz
- Postdoc and Marie-Curie Individual Fellow at University of Cambridge (2017-2019)
- Assistant Professor at Eindhoven University of Technology (2019-2021)
- Current: Assistant Professor at Graz University of Technology
Research Interests
Peharz's research spans multiple areas of artificial intelligence with a focus on making probabilistic reasoning both theoretically sound and practically efficient. His work addresses fundamental challenges in tractable probabilistic inference and learning, probabilistic circuits as a unified framework for deep generative models, Bayesian causal inference, and neurosymbolic AI combining sub-symbolic and symbolic approaches. His research has applications in cybersecurity, healthcare, and energy systems.
Research Projects
- VENTUS (2024-present): Physics-informed, probabilistic and causal machine learning for wind energy systems
- NEO DNA (2023-present): DNA-based data storage systems using computer vision and probabilistic ML
- VanillaFlow (2023-present): AI-guided development of novel vanillin-based molecules for redox flow batteries
- Bilateral AI: Cluster of Excellence focused on Broad AI combining sub-symbolic and symbolic AI approaches
Awards and Recognition
- Finalist for TUG's Excellent Teaching Award (2023) for all 3 of his courses
- Marie-Curie Individual Fellow at University of Cambridge
Academic Service
Peharz is actively involved in the academic community through conference organization and reviewing:
- Area Chair: UAI (2022), ECML/PKDD (2022)
- Senior Committee Member: UAI (2021), IJCAI (2019, 2020)
- Reviewer for major conferences including ICML, NeurIPS, AAAI, IJCAI-ECAI
Teaching and Mentorship
Peharz supervises multiple PhD students working on diverse projects at the intersection of machine learning, causality, and neurosymbolic AI. His current advisees include Sepideh Adamiat, Irina Dobrianski, Johannes Exenberger, Giacomo Di Gobbi, Tim d'Hondt, Christian Toth, and Thomas Wedenig. Previous students include Alvaro Correia, Martin Trapp, and David Montalvan.
