Tutor: Dr George Giannakopoulos
🏛 ️INSANE Group - NCSR “Demokritos” (GR)
🏛 SciFY PNPC (GR)
🏛 ahedd DIH S.A. (GR)
✉ ggianna<at>iit.demokritos.gr
Short CV
George Giannakopoulos, PhD, is an Artificial Intelligence (AI) researcher at NCSR Demokritos (Greece), as well as co-founder, and CEO of SciFY, a multiply awarded not-for-profit Company on applying Artificial Intelligence to everyday life for the common good. He leads the Intelligent Science and Engineering (INSANE) Group of the Institute of Informatics and Telecommunications of NCSR Demokritos, working on the interaction between AI, science and engineering. He is also co-founder and scientific officer of the ahedd DIH S.A., a BDVA Platinum i-Space spin-off of NCSR “Demokritos”. Dr Giannakopoulos has 20 years of AI-related experience on international and national research and innovation, but also numerous industrial projects. He has more than 25 years of IT consulting and software engineering expertise in domains such as AI, Machine Learning and Natural Language Processing, Bio-medical and chemical informatics, the Semantic Web and others. He has also co-authored more than 100 scientific publications. He is a member of the Hellenic Artificial Intelligence Society (EETN) and a member of the European Chapter of the Association of Computational Linguistics (ACL). He has also contributed at a national policy-making level through the Sectorial Scientific Council on Data Policy and A.I. of the National Council for Research, Technology and Innovation. He has offered tens of presentations and tutorials related to AI and its application, to varying audiences across disciplines. These dissemination actions have been complemented by the Greek book “AI: A subtle demystification” (Greek title: “Τεχνητή νοημοσύνη – Μια διακριτική απομυθοποίηση”), published in 2021 by Ropi Publications.
Tutor: Dr Aris Kosmopoulos
🏛 SciFY PNPC (GR)
✉ akosmo<at>scify.org
Short CV
Aris Kosmopoulos, PhD, has more than 15 years experience in Artificial Intelligence Research, focusing on large scale machine learning, natural language processing and evaluation. He has worked in several research EU projects related to AI and has more than 15 years of experience in software development related to AI products. Dr Kosmopoulos has received his BSc, MSc and Phd from the department of Computer Science of AUEB in collaboration with NCSR Demokritos and he brings more than10 years of research and industrial experience as an AI Expert in SciFY and NCSR “Demokritos”, building real world systems and models in domains such as biomedical informatics, natural sciences and engineering, natural language processing, text and data mining, legal informatics and more. He has also contributed to numerous awarded projects, such as the IRCAI Top 100 project “Pioneers of AI in Greece”, “AI Pioneers in Natural Sciences” training for new scientists,“People with Disabilities Pioneers in AI”, building capacity for more inclusive and democratized AI landscape.
Tutorial Description
AI has become an invaluable tool across disciplines. With proven value throughout the scientific research pipeline and numerous engineering challenges, AI researchers have become a necessary addition for many interdisciplinary teams. However, being able to model problems from other disciplines appropriately is far from a trivial task: it requires knowledge, agility and an understanding of the unique challenges these settings offer. This tutorial discusses the interaction between Artificial Intelligence and the scientific and engineering research pipelines. We will overview a set of inter-disciplinary scientific problems using AI, originating from domains including indicatively energy, environment, nuclear physics, material design, mathematics, structural mechanics, bio-medicine, natural disasters resilience and more. We will describe the very interesting AI problems that arise from these scientific domains and we will discuss various use cases and the unique viewpoints they offer, by providing challenges related to: data-lean and frugal settings, cost-aware and multi-fidelity data generation, interpretable model building and more.
During the tutorial we will outline state-of-the-art challenges related to
- AI-guided experimental design,
- AI-guided data generation,
- efficient design space exploration,
- explainable science and engineering models,
- knowledge-informed / knowledge-infused machine learning (ML),
- knowledge-guided and validated GenAI and other interesting AI approaches.
Upon completion, the participants will be able to
- enumerate recent challenges and novel viewpoints over AI/ML problems, inspired by other disciplines and their real-world needs;
- map AI/ML approaches to deal with problems in natural sciences and engineering;
- enumerate approaches that deal with these new challenges and recent requirements of systems, going beyond brute force deep learning models and data- and computation-hungry GenAI approaches.
Intended Audience & Prerequisites
Intended audience
- Graduate and post-graduate students of informatics and related topics
- Graduate and post-graduate students of engineering and natural sciences, with an interest in AI
- AI scientists and engineers with cross-disciplinary interests
Prerequisites
- Understanding of graduate level mathematics
- Very basic understanding of AI terms and machine learning
Duration
2-3 hours.