https://uniba.sk/ Najnovšie novinky sk https://uniba.sk/fileadmin/tt_news_article.gif https://uniba.sk/ Najnovšie novinky TYPO3 - get.content.right http://blogs.law.harvard.edu/tech/rss Thu, 31 Dec 2026 00:00:00 +0100 http://https://uniba.sk/en/comenius-university-news/news-detail/back_to_page/univerzita-komenskeho/article/internationalisation-can-attract-and-retain-top-talent-1/ At the end of 2024, an Audit Expert Panel visited Comenius University to evaluate the institution's... [FaF] Titulka - úzke novinky EN Thu, 20 Mar 2025 08:17:00 +0100 http://https://uniba.sk/en/comenius-university-news/news-detail/back_to_page/univerzita-komenskeho/article/najviac-excelentnej-vedy-na-slovensku-vznika-na-uk/ Bratislava, 10 March, 2025 – Comenius University Bratislava (CU) produces the highest-quality... [FaF] Titulka - úzke novinky EN Thu, 20 Mar 2025 08:14:00 +0100 https://uniba.sk/ Monday 24.3.2025 at 13:10, Lecture room I/9 Ján Pastorek:
Graph isomorphism, asymmetric graphs and partial symmetries Abstract:
Graphs, composed of vertices connected by edges, model many complex systems encountered in science and engineering. Although almost all graphs are asymmetric—that is, they admit only the trivial automorphism—some graphs appear intuitively symmetric even though they possess no nontrivial automorphisms. The Graph Isomorphism Problem (GI), which asks whether two graphs are structurally identical, and the Graph Asymmetry Problem (GA), which determines whether a graph is asymmetric, are two prominent unresolved problems in theoretical computer science. Their computational complexity remains open; neither problem is known to belong to class P nor to be NP-complete. Moreover, it is unclear whether GI can be reduced to GA or what the precise relationship between symmetry and asymmetry in graphs might be. In his plenary talk at ICM 2018, Babai suggested that new insights into GI and GA might arise from a deeper understanding of the local structure of graphs. Local structure of graphs can be studied using partial automorphisms—namely, isomorphisms between induced subgraphs. In our work, we define the symmetry level of a graph as the ratio of the largest rank (i.e., the size of the domain) of a nontrivial partial automorphism to the order of the graph. We further refine theoretical bounds on this measure and identify extremal examples through the use of parallel algorithms implemented in Julia.Fotku prikladam do prilohy. More information ]]> [FMFI] Prednášky a semináre [FMFI] Kalendár Damas.Gruska@fmph.uniba.sk Wed, 19 Mar 2025 08:06:00 +0100 https://uniba.sk/ Thursday 20.3.2025 at 9:50, Lecture room M 213 Martin Mačaj:
r-Regular families of permutations and 1-factorizations of bipartite multigraphs Abstract:
It is well known that a k-valent bipartite multigraph has a 1-factorization. Motivated by the notion of r-regular families of permutations, introduced by R. Jajcay and G. Jones, we will discuss what happens when we require 1-factors to be mutually distinct (in the uderlying simple graph). This is a joint work with P. Kollar.  More information]]> [FMFI] Prednášky a semináre [FMFI] Kalendár Martin.Skoviera@fmph.uniba.sk Tue, 18 Mar 2025 11:46:00 +0100 https://uniba.sk/ From 26 February to 2 March 2025, Comenius University Bratislava (CU) participated in the... "A total of around 500 students expressed interest in studying at our university, confirming our analyses, which identified the Central Asian region as a priority for recruiting international students," said Martin Považan, Head of the International Relations Office at CU. In addition to participating in the fairs, the delegation also held productive meetings with the ambassadors of the Slovak Republic in Uzbekistan and Kazakhstan and established contacts with foreign universities to explore cooperation opportunities. These activities contribute to strengthening CU's international position and open new opportunities for recruiting talented students from abroad and developing strategic partnerships. CU's participation in the fairs was supported by the Recovery and Resilience Plan of the Slovak Republic under the call for "Support for Projects Promoting Universities Abroad." Financed by European union NextGenerationEU]]> [RUK] Titulka - široké novinky [RUK] Spravodajský portál pr@uniba.sk Thu, 20 Mar 2025 11:21:00 +0100 https://uniba.sk/ Krisztián Fenyves, a PhD candidate in Patristic Studies (Ancient Christian Literature and History)... SAIA website Krisztián Fenyves describes his 10-month research stay at RKCMBF UK within the National Scholarship Programme, his excellent cooperation with assoc. prof. Jozef Jančovič, under whose guidance the PhD student worked in Slovakia, and the diverse opportunities not only in the field of his professional studies and work, but also in the field of personal development and getting to know Slovakia. Krisztián Fenyves's very well elaborated paper is available on the SAIA website: Faith, Research, and Discovery: Exploring Imago Dei in Bratislava.]]> [RKCMBF] Titulka - široké novinky Mon, 17 Mar 2025 17:24:00 +0100 https://uniba.sk/ Thursday 20.3.2025 at 13:00, Lecture room F1 108 Ing. Jaroslav Tóbik, PhD. (Institute of Electrical Engineering SAS):
Ferrotorroidal order by spontaneous symmetry breaking
Abstract:
The development of technology makes it possible to fabricate artificial structures on a sub-micrometer scale with a good definition of building elements. The periodic repetition of the patterned elements defines a crystal. Thus, technology allows us to build artificial materials with properties that are not observed in natural materials. In the field of applied magnetism, there is vivid research in artificial structurally periodic arrays of magnetic elements. However, these structures usually have a non-periodic magnetic pattern. There are numerous reasons for difficulties in achieving magnetic order. The main obstacle is the presence of many local minima and large energy barriers among them. The energy barriers are usually much higher than the achievable thermal energy. Therefore, thermally activated magnetization flipping is possible only during the growth process when the thickness of the magnetic material is about 1 nm. Then, the magnetic structure freezes. The magnetic state ma- nipulation afterwards is problematic, especially in cases when the ground state has no net magnetization, as in the case of ferrotoroic order. To tackle this problem we suggested new demagnetization protocol. It is based on magnetizing the structures into out-of-plane direction. Naive expectation is that the tilt of the magnetization into out-of-plane direction will only renormalize energy landscape. More detailed analysis shows, that by this procedure the ferrotoroic state is the only stable state in the basic unit cell for certain interval of external magnetic field. Thus appearance of stable and metastable magnetic states in some structures can be described as a serie of symmetry breaking processes. This theory was verified by experiments with rather convincing results. ]]> [FMFI] Prednášky a semináre [FMFI] Kalendár Roman.Martonak@fmph.uniba.sk Mon, 17 Mar 2025 15:38:00 +0100 https://uniba.sk/ Thursday 20.3.2025 at 14:00, Lecture room M 223 Anna Hudertmark (RPTU Kaiserslautern-Landau):
Modeling of carotid artery hemodynamics towards a CFD database for stroke risk assessment
Abstract:
In medicine, the risk assessment of strokes caused by pathological irregularities in the bifurcation of the carotid artery plays a major role in the patient's optimal treatment decision. This decision-making process can be greatly enhanced by detailed flow field data obtained from simulations using patient-specific arterial morphologies. We present our numerical modeling approach to create a CFD database that can guide stroke risk assessment, including hemodynamic shear risk metrics. We investigate the multi-directional behavior of wall shear stress (WSS) based on specially constructed longitudinal tangent vectors using a center-line projection approach and present the utility of longitudinal WSS evaluation for detecting opposing or transverse WSS in slow patient-specific flow. In our fluid-structure interaction model, we generalize the assumption of linear elasticity by a strain-dependent elastic modulus that is consistent with laboratory measurements of compliance and distensibility. Finally, we present the established working pipeline to generate CFD data for a large number of patient-specific geometries. Created hemodynamic datasets can be used for visual atlases and comparative analysis or explored by neural networks. ]]> [FMFI] Prednášky a semináre [FMFI] Kalendár Pavol.Quittner@fmph.uniba.sk Mon, 17 Mar 2025 13:46:00 +0100 https://uniba.sk/ Wednesday 19.3.2025 at 14:00, Lecture room F1/364 Mgr. Matúš Kosťov:
Fast Detector Simulation and Novel Reconstruction Techniques in the ATLAS experiment - obtaining scales and resolution of large-R jets for AtlFast3 simulation Abstract:
In 2022, a new operational period of the LHC accelerator in CERN Run 3 started. The expected yield of the integrated luminosity is around 370〖fb〗^(-1), which is approximately 2.5 times more than the yield of the Run 2 period (139〖fb〗^(-1)). It is clear that with the increasing amount of data, the precise but CPU-heavy, full simulations will not be able to keep up in the future, especially after the High-Luminosity upgrade. Therefore, ATLAS is developing and improving fast simulations. One such simulation is AtlFast3, which should take over a portion of simulated events in future. Furthermore, a novel technique called Unified Flow Objects was developed, which reconstructs objects by combining information from tracker detectors and calorimeters to provide better resolution. However, these methods need to be calibrated and validated so that they can be employed in the Run 3 analyses.]]> [FMFI] Prednášky a semináre [FMFI] Kalendár Jaroslav.Stanicek@fmph.uniba.sk Thu, 13 Mar 2025 14:31:00 +0100 https://uniba.sk/ Monday 17.3.2025 at 13:10, Lecture room I/9 Filip Kerák:
Training Sparse Neural Networks Abstract:
Modern neural networks easily contain billions of parameters, which leads to high computation requirements during both the training and inference phases. An astronomical amount of money and resources is used to create models that, on one hand, achieve state-of-the-art results in all areas but are not widely feasible. One way to tackle this problem is to reduce the number of parameters in those networks. Best improvement can be achieved by introducing sparsity as the key part of the training. Different sparse training approaches for pruning and growing networks are being used, each trading performance for accuracy and vice versa. What is the "golden mean"? More information ]]> [FMFI] Prednášky a semináre [FMFI] Kalendár Damas.Gruska@fmph.uniba.sk Thu, 13 Mar 2025 08:25:00 +0100