Keynote Speakers

Mario Guarracino

University of Cassino and Southern Lazio, Italy

Topics

Graph Embedding, Network Analysis, Statistical Learning.

Biography

Mario Guarracino is a data scientist specializing in statistical learning and artificial intelligence, with a focus on statistical graph embedding. He is an Associate Professor of Statistics at the University of Cassino and Southern Lazio.

He earned his PhD in Mathematics with a dissertation on projection techniques for parallel sparse linear algebra, after completing an MSc in Applied Mathematics. His postdoctoral research at the National Research Council of Italy focused on developing low-cost, high-performance supercomputers for scientific computing. Since 2005, he has collaborated with the Center for Applied Optimization at the University of Florida, and since 2014, he has served as Chief Research Fellow at the Laboratory of Algorithms and Technologies for Network Analysis at the Higher School of Economics. From 2012 to 2019, he directed the Computational Data Science Laboratory (formerly the Laboratory for Genomics, Transcriptomics, and Proteomics) at the National Research Council.

His research aims to develop novel algorithms, methods, and software tools based on mathematical programming, artificial intelligence, and statistical learning theory, with applications to network analysis. He applies this expertise to address complex scientific and economic problems, helping researchers and practitioners integrate and interpret large-scale, heterogeneous data.

Talk

Embedding Nodes and Whole Graphs: A Statistical View on Learning from Networked Data

Graphs offer a convenient framework for representing complex systems, capturing interactions at multiple levels—from individual nodes to substructures and entire networks. In fields such as biomedicine, bioinformatics, economics, and finance, the ability to encode and analyze these relational structures is critical for inference, predictive modeling and exploratory analysis.

This talk surveys modern embedding techniques that map nodes and whole graphs into low-dimensional vector spaces, enabling the use of classical statistical learning algorithms on inherently non-Euclidean data. We discuss a variety of methods, including graph kernels, matrix factorization, and deep learning approaches, focusing on their ability to preserve structural information and support tasks such as classification, clustering, and representation learning.

Emphasis is placed on the statistical modeling principles underlying these techniques, including trade-offs between expressivity, interpretability, and computational efficiency. Through examples in biomedical networks, molecular graphs, and economic interaction data, we illustrate how embedding choices impact downstream performance and guide the selection of appropriate methods for real-world problems.

Panos Pardalos

University of Florida, USA

Topics

Global Optimization, Mathematical Modeling, Energy Systems, Financial applications, and Data Sciences

Biography

Panos Pardalos was born in Drosato (Mezilo) Argitheas in 1954 and graduated from Athens University (Department of Mathematics). He received his PhD (Computer and Information Sciences) from the University of Minnesota. He is a Distinguished Emeritus Professor in the Department of Industrial and Systems Engineering at the University of Florida, and an affiliated faculty of Biomedical Engineering and Computer Science & Information & Engineering departments.

Panos Pardalos is a world-renowned leader in Global Optimization, Mathematical Modeling, Energy Systems, Financial applications, and Data Sciences. He is a Fellow of AAAS, AAIA, AIMBE, EUROPT, and INFORMS and was awarded the 2013 Constantin Caratheodory Prize of the International Society of Global Optimization. In addition, Panos Pardalos has been awarded the 2013 EURO Gold Medal prize bestowed by the Association for European Operational Research Societies. This medal is the preeminent European award given to Operations Research (OR) professionals for “scientific contributions that stand the test of time.”

Panos Pardalos has been awarded a prestigious Humboldt Research Award (2018-2019). The Humboldt Research Award is granted in recognition of a researcher’s entire achievements to date – fundamental discoveries, new theories, insights that have had significant impact on their discipline.

Panos Pardalos is also a Member of several Academies of Sciences, and he holds several honorary PhD degrees and affiliations. He is the Founding Editor of Optimization Letters, Energy Systems, and Co-Founder of the International Journal of Global Optimization, Computational Management Science, and Springer Nature Operations Research Forum. He has published over 600 journal papers, and edited/authored over 200 books. He is one of the most cited authors and has graduated 71 PhD students so far. Details can be found in www.ise.ufl.edu/pardalos

Panos Pardalos has lectured and given invited keynote addresses worldwide in countries including Austria, Australia, Azerbaijan, Belgium, Brazil, Canada, Chile, China, Czech Republic, Denmark, Egypt, England, France, Finland, Germany, Greece, Holland, Hong Kong, Hungary, Iceland, Ireland, Italy, Japan, Lithuania, Mexico, Mongolia, Montenegro, New Zealand, Norway, Peru, Portugal, Russia, South Korea, Singapore, Serbia, South Africa, Spain, Sweden, Switzerland, Taiwan, Turkey, Ukraine, United Arab Emirates, and the USA.

Talk

Introduction to data analytics for networks – a historical perspective and major advances

Data analytics for networks involves the use of advanced techniques and tools to extract insights and knowledge from large and complex datasets generated by network devices, applications, and services. This process involves collecting, storing, processing, and analyzing large amounts of data to identify patterns, trends, and anomalies that can provide valuable information for network operators. By leveraging data analytics, network researchers can make informed decisions about network planning, capacity management, service delivery, and customer experience. Additionally, data analytics can help network operators to detect and respond to security threats and attacks, by analyzing network traffic, identifying abnormal behavior, and detecting potential vulnerabilities. Overall, data analytics is a critical component of massive networks, enabling network researchers to extract valuable insights from massive datasets and improve network performance, efficiency, and security. Many optimization problems related to the analysis of networks are large-scale and non-convex.

Panos Pardalos, www.ise.ufl.edu/pardalos

The other Keynote Speakers

Yuki Asano, University of Technology Nuremberg, Germany
Pierre Baldi, University of California, Irvine, CA,USA
Lucas Beyer, OpenAI, Zürich, Switzerland
Wessel Bruinsma, Microsoft Research Amsterdam, The Netherlands
Sam Buchanan, Toyota Technological Institute at Chicago, USA
Floris Geerts, University of Antwerp, Belgium
Sven Giesselbach, Fraunhofer Institute – IAIS, Germany & Telecom Systems
Vicky Kalogeiton, Ecole Polytechnique Paris, France
Thomas Kipf, Google DeepMind, USA
Yi Ma, University of California, Berkeley, USA
Qing Qu, University of Michigan, USA
Abigail See, Google DeepMind, London, UK
Michal Valko, Meta Paris, France
Yaodong Yu, OpenAI, USA
Daniel Zügner, Microsoft Research Berlin, Germany

are those of International Artificial Intelligence Summer School – IAISS 2025, September 21-25, 2025, Riva del Sole Resort & SPA, Castiglione della Pescaia (Grosseto), Tuscany, Italy a Residential Summer School in Artificial Intelligence & Generative Artificial Intelligence for Science and Engineering, which is simultaneously in the same conference venue (Riva del Sole Resort & SPA).