Title: Integrating Data-and Knowledge-Driven Approaches to Automated Scientific Modelling
Speaker: Sašo Džeroski, Head of the Department of Knowledge technologies, Jozef Stefan Institute, Ljubljana, Slovenia
Organized by Unidad Asociada CSIC-CESGA “Optimización computacional en biología de sistemas”.
Friday 15th decembre, 2023, 11:00h.
Salón de Actos Misión Biológica de Galicia, Instituto de Investigaciones Agrobiológicas – CSIC Avda. de Vigo, s/n (Campus Vida) Santiago de Compostela.
In knowledge-driven modelling, an expert derives a model based on their knowledge of the domain studied: Both the structure and the parameters of the model are derived by the expert from knowledge about the entities and processes in the modelled system. In data-driven modelling, many model structures are considered in a trial-and-error fashion, their parameters are fit to data, and a complete model is returned: This is typically a black-box process that does not take into account domain knowledge. Explainable scientific models need to be expressed in formalisms accessible to humans and learned through approaches that integrate data-driven and knowledge-driven modeling and use both data and domain knowledge.
The talk will discuss approaches to integrating data-driven and knowledge-driven construction of scientific models. Different formalisms for representing models and domain knowledge will be discussed, including process-based models and context-free grammars. The talk will conclude with a discussion of recent approaches that rely on the use of probabilistic context-free grammars and other generative models for equation discovery.
Speaker Google scholar profile – https://scholar.google.com/citations?hl=en&user=_aIV-aEAAAAJ
Sašo Džeroski is a scientific councilor (senior researcher) at JSI, full professor at the Jozef Stefan International Postgraduate School, and visiting professor at the European Space Agency. He is a fellow of EurAI, the European Association of AI, in recognition of his “Pioneering Work in the field of AI”. He is a member of the Macedonian Academy of Sciences and Arts and a member of Academia Europea.
His research covers different facets of Artificial Intelligence, focusing on machine learning and computational scientific discovery. His group has developed machine learning methods that learn explainable models from complex data in the presence of domain knowledge: These include methods for multi-target prediction, semi-supervised and relational learning, and learning from data streams. It has also developed methods for computational scientific discovery, including methods that learn explainable models of dynamical systems. The developed methods, released in open-source software, have been used to solve important problems in science and society, including agriculture and environmental sciences, medicine and life sciences, physics and material sciences, and space operations/Earth observation.
The work of Professor Džeroski has been extensively published in more than 250 journal papers and more than 300 conference papers. It is highly cited, with over 24000 citations and an h-index of 73 (in the GoogleScholar database). Prof.Džeroskiis recognized as the best Computer Scientist in Slovenia – according to the ranking by Research.com. He has supervised 30 PhD students that have successfully defended their theses, as well as 15 PostDoc fellows, now active in both academia and industry in many countries across three continents.