On Thursday 18 March, starting at 2 p.m., we will make presentations of further preLABs active in the DigiWorld PRA:

PreLAB: Autonomatic Scientific Discovery Lab

(delivered in Polish)

Speakers:

• Dr Stanisław Jastrzębski, scientific director at Molecule.one, Institute of Computer Science and Computer Mathematics, JU
• Dr Bartosz Zieliński, Institute of Computer Science and Computer Mathematics, JU

Title: overview of the lab and interpretability of models applied to medicine

Abstract: What if a machine could discover a new physical matter or improve our understanding of disease? Due to rapid advances in artificial intelligence and hardware to collect data, such questions are no longer hypothetical. 

Our lab focuses on developing relevant generic artificial intelligence methods in two application areas: medicine and chemistry. These two domains are characterized by access to troves of data that have become challenging to analyze by humans. 

In the first part of the talk, we will give a short introduction to the field and describe the motivation behind our lab. In the second part of the talk, we will briefly discuss some of our recent application works in medicine, with a focus on interpretability.

Speakers:

Dr Stanisław Jastrzębski works as Chief Scientific Officer at Molecule.one and as academic staff at the Jagiellonian University. He defended his doctoral thesis at the Jagiellonian University, supervised by professor Jacek Tabor and professor Amos Storkey (University of Edinburgh). He worked as Postdoctoral Fellow at New York University. He is actively involved in the goings-on of the machine learning community as the host of conferences, member of the MLinPL, area chair and reviewer. He believes that in the future artificial intelligence methods will actively help in scientific discoveries. In his research work, he pursues this vision by means of developing applications in drug discovery and developing new methods for training neural networks.

Dr Bartosz Zieliński graduated in Computer Science from the Faculty of Mathematics and Computer Science, JU. He earned his doctorate in technology, in computer science, in 2012 from the Institute of Fundamental Technological Research, Polish Academy of Science, and has since worked as assistant professor at the Chair in Applied Computer Science, JU. He has authored or co-authored numerous articles in journals listed in Journal Citation Reports and publications at conferences NeurIPS, IJCAI and WACV. Currently he is a member of the InfoTech team in the project DNP Team-Net, dealing with biologically-inspired artificial neural networks. He also works as Lead Data Scientist at the company Ardigen S.A., where he and his team implement models of deep learning to problems of biomedical imaging.

PreLAB: InfoLab

Speakers:

• Dr hab. Jakub Mielczarek, Faculty of Physics, Astronomy and Applied Computer Science
• Dr Małgorzata Budzanowska-Drzewiecka, Institute of Economics, Finance and Management
• Dr Bartosz Lisowski, Chair of Physiology
• Dr hab. Kamilla Małek, prof. UJ, Faculty of Chemistry
• Dr Kamil Korzekwa, Faculty of Physics, Astronomy and Applied Computer Science
• Dr Tadeusz Pałasz, Faculty of Physics, Astronomy and Applied Computer Science

The presentation will introduce InfoLab: Jagiellonian Center of Biomedical Imaging. The InfoLab’s thematic scope concentrates around information: its processing, storage, protection and social impact. The research area of the InfoLab covers a broad range of problems which deal with the classical and quantum information theory, both in terms of of physical data media and on the abstract level of its processing and role in the contemporary world. The paramount goal formulated by the InfoLab for itself is to develop information processing and exchange methods based on quantum phenomena, and to implement them in different areas of science (physics, chemistry, pharmacy, biology, medicine, social sciences), technology, business and social life. In particular, it pertains to the use of quantum algorithms (including machine learning quantum algorithms) in processing large datasets and solving complex computational problems, information exchange security through quantum cryptography, as well as collecting and storing information using quantum phenomena. 

Presentation plan:

Dr hab. Jakub Mielczarek: ‘InfoLab’

Dr Małgorzata Budzanowska-Drzewiecka: ‘Dilemmas of consumer reception of market information. Sender hints in eWOM communication processing.’

Dr Bartosz Lisowski: ‘How do Mycetozoa learn (and what it costs) or on the mechanisms of non-neural memory’

Dr hab. Kamilla Małek, prof. UJ,

Dr Kamil Korzekwa: ‘Quantum information’

Dr Tadeusz Pałasz: ‘The qutools kit for quantum physics’

Time: 18 March 2021 (Thursday, 2 p.m. GMT+1)

Access: through MS Teams (code: tcmk6fo) or the link https://tinyurl.com/SeminariumDigiWorld

For more information go to the DigiWorld PRA website and check the Seminars tab, or connect to Teams.