CLOUD COMPUTING SEMINAR SERIES
Speaker : Prof. Radu Prodan, University of Klagenfurt, Austria.
Title :”Think Broader, Run Local: Towards Neuro-Symbolic AI at the Edge”
Date & Time : December 17, 2024, 11:30 AM
Venue : # 419, CDS 4th Floor.
ABSTRACT
The presentation starts by reviewing the convergence of two historically disconnected AI branches: symbolic AI of explicit deductive reasoning owning mathematical rigor and interpretability, and connectionist AI of implicit inductive reasoning lacking readability and explainability.
Afterward, it makes a case of neuro-symbolic data processing in knowledge graph representation approached by the Graph-Massivizer Horizon Europe project and applied for anomaly prediction in data centers using graph neural networks. Machine learning-driven graph sampling algorithms support its training and inference on resource-constrained Edge devices. The presentation concludes with an outlook towards a broad AI vision.
BIOGRAPHY
Radu Prodan received the PhD degree from the Vienna University of Technology in 2004. He is a professor in distributed systems with the Institute of Information Technology, University of Klagenfurt, Austria. Previously, he was an associate professor with the University of Innsbruck, Austria. His research interests are performance optimization, and resource management tools for parallel and distributed systems. He participated in numerous projects and coordinated the European Union projects ARTICONF and Graph-Massivizer. He has co-authored more than 250 publications and received three IEEE best paper awards.
Host Faculty: Prof. Yogesh Simmhan
ALL ARE WELCOME
