This lecture explores how machine learning can help identify, categorize, and forecast events of interest when working with limited data and imprecise labels. By combining structured and unstructured multimodal information, the presented approaches model different types of interactions within temporal multilayer networks.
The findings show that learning frameworks capable of integrating information from diverse sources — even when these sources vary in quality and resolution — can significantly improve informed decision-making in complex real-world environments.
The lecture will be delivered by Zoran Obradovic, Distinguished Professor and Center Director at Temple University, Academician of Academia Europaea, and Foreign Academician of the Serbian Academy of Sciences and Arts. His research focuses on data science and complex networks in decision support systems, with applications in healthcare management, power systems, earth sciences, and social sciences.
The lecture will be held in English. Admission is free.