Once a week during the academic year, an invited speaker (from abroad, as well as from the Czech Republic) talks about his or her scientific work. Colloquium takes place at the Faculty of Informatics and is open to the scientific community. Lecture dates can be found in the programme. Tuesday 14.00 - 15.00, D2, FI MU, Botanická 68a

Colloquia programme with abstracts for the Autumn 2015 semester

29. 9. 2015
RNDr. Vojtěch Řehák, Ph.D., FI MU
Patrolovací hry
Abstract: Patrolovací hry jsou zavedeným konceptem pro modelování a plánování ostrahy chráněných objektů. Obránci se pochůzkou po chráněných objektech snaží zabránit útočníkům v provedení úspěšného útoku. Po zkušenostech s organizovaným zločinem se v těchto hrách předpokládá, že útočníci vidí nejen aktuální pozici obránců, ale znají i jejich strategii. Jedinou nadějí obránců je pak aplikace pravděpodobnostních strategií. Automatická tvorba obranných strategií byla úspěšně použita v mnoha praktických případech o velkém počtu chráněných objektů i obránců.

My se zaměříme na základy této problematiky. Umíme najít optimální strategii? Existuje vůbec? Na pár drobných příkladech si ukážeme záludnosti patrolovacích her a nebude chybět ani vyvrácení přirozených, leč mylných předpokladů použitých v některých publikovaných výsledcích.

6. 10. 2015
prof. Cristian S. Calude, The University of Auckland
Pluses and minuses of "big data" approach
The deluge of spurious correlations in Big Data
Abstract: Very large databases are a major opportunity for science and data analytics is a remarkable new field of investigation in computer science. However, the effectiveness of these tools seems to encourage an aggressive ``philosophy’' against the scientific method as developed along history According to this view, computer discovered correlations should replace understanding and guide prediction and action. The ``End of Science'' is proudly proclaimed.

There is no need to give scientific meaning to phenomena as ``with enough data, the numbers speak for themselves''. We show, using the classical Ramsey theorem, that this ``philosophy'' is radically wrong. Specifically, we prove that, exactly because of their very large size, databases have to contain arbitrary correlations, most of them spurious. These correlations appear only due to the size not to the nature of the data. The scientific method can be enriched by computer mining over immense databases, but surely not replaced by it.

13. 10. 2015
doc. JUDr. Radim Polčák, Ph.D., PrF MU
Kybernetická bezpečnost pro ježky
Abstract: Struktura zákona o kybernetické bezpečnosti obsahuje řadu partikulárních kontradikcí a problémových momentů. Jedná se ale každopádně o první komplexní úpravu tohoto typu, která v demokratickém právním státě úspěšně prošla legislativním procesem. Důvodem politického úspěchu nově vytvořeného právního fenoménu kybernetické bezpečnosti je především implicitní obsah hodnot, které nemají kontradiktorní povahu. Zatímco jsou jiné součásti bezpečnostního práva přirozeně hodnotově kontradiktorní (tj. jde o konflikt bezpečnosti a soukromí, bezpečnosti a vlastnictví nebo třeba bezpečnosti a autonomie vůle), není české právní řešení postaveno na nutnosti proporcionálně vážit protikladné zájmy. Vedle této doposud nepublikované problematiky představí příspěvek též vybrané problémové momenty nové české úpravy, přičemž pozornost bude věnována především otázce compliance a právní odpovědnosti jednotlivce.
13. 10. 2015 v 15:15h
Prof. Thomas A. Henzinger, Ph.D., Dr.h.c., IST Austria
Reactive Systems: A Powerful Paradigm for Modeling and Analysis from Engineering to Biology
Abstract: A reactive system is a dynamic system that evolves in time by reacting to external events. Hardware components and software processes are reactive systems that interact with each other and with their physical environment. Computer science has developed powerful models, theories, algorithms, and tools for analyzing and predicting the behavior of reactive systems. These techniques are based on mathematical logic, theory of computation, programming languages, and game theory. They were originally developed to let us build a more dependable computer infrastructure, but their utility transcends computer science. For example, both an aircraft and a living organism are complex reactive systems. Our understanding and the design of such systems can benefit greatly from reactive modeling and analysis techniques such as execution, composition, and abstraction.
20. 10. 2015
assoc. prof. Rong Qu, Ph.D., School of Computer Science, University of Nothingham
Recent Developments on Real World Optimization Problems
Abstract: Combinatorial optimisation problems appear in many real world applications in management, organisation and business sectors. The talk presents some recent research developments on hyper-heuristic for various optimisation problems. A general hyper-heuristic framework is then presented for solving educational timetabling problems. The general framework adaptively searches and calls low level constructive heuristics. It is easy to construct, and has been extended and utilised by researchers on different applications. Several research issues are also discussed to deal with fundamentals of this framework. Research developments of using a range of hybrid techniques for other scheduling and optimisation problems will also be briefly overviewed.
Biography: Dr. Rong Qu is an Associated Professor at the University of Nottingham. Dr. Qu's main research interests include the modelling and optimisation algorithms for scheduling and optimisation algorithms in transport scheduling in logistics, personnel scheduling, telecommunication network routing, portfolio optimisation, and timetabling problems, etc. by using evolutionary algorithms, mathematical programming, constraint programming in operational research and artificial intelligence, and hybridisations of these techniques.

Dr. Qu has published more than 50 peer-refereed papers at international journals since 2000. Among these several have been awarded the Top Cited Paper at leading OR journals, and ranked the top 1%, or top 10% cited papers by ISI Essential Science Indicators. Dr. Qu has been the program chair of several symposium, workshops and special sessions and guest editor of a special issue at Journal of Scheduling. She is the chair of Task Force of Hyper-heuristics at IEEE Computational Intelligence Society, and elected as an IEEE Senior Member in 2012. She has served the programme committee of more than 50 international conferences.

27. 10. 2015
doc. Eva Dokladalova, ESIEE Paris
EFFICIENT ARCHITECTURES AND ALGORITHMS FOR EMBEDDED VISION SYSTEMS
Abstract: In this talk we develop three main axes i) design of efficient hardware architectures, ii) computational efficient algorithms targeted for embedded vision systems and iii) hardware support for self-aware computing.

We will introduce recent advances within the unifying framework of mathematical morphology. We propose a first morphological processor with arbitrarily large neighborhoods. It allows to obtain previously unachieved performances for serially composed morphological filters, geodesical and conditional operators. The cited processor is based on a novel algorithm formulation of morphological dilation.

Finally, the applicative domain will be illustrated in scene understanding context for self aware embedded computing.

3. 11. 2015
Dr. Matteo Olivieri, Ph.D., University of Calabria
Open problems in economic logic: is there anything that computer scientists can learn from business scientists?
Abstrakt: Key challenges in modern computer science include neural and quantum computing, as well as the attempt to design neurally-inspired processes to replicate the cognitive functions of the brain. Is there anything that computer scientists can learn from business scientists? After all, the ability to handle complexity and detect patterns out of apparent randomness lies at the core of evolution of both disciplines.
Životopis: Matteo Olivieri received a Ph.D. in Business Sciences from the University of Calabria, Italy, with a concentration in Corporate Social Responsibility. He is a post-doctoral fellow in economic, social- and environmental sustainability at the same university. He has extensive teaching experience at an international level, including Strategy and Leadership in the Masaryk University’s Service Science, Management and Engineering master program. His research spans different areas of management, and much of it occurs at the intersection of business sciences and cognitive sciences. He investigates questions like the foundation of value and the economic, social and environmental impacts of corporate decisions.
10. 11. 2015
prof. Dr. Ing. Zdeněk Hanzálek, Fakulta elektrotechnická ČVUT
Scheduling algorithms for time-trigged systems
Abstrakt: Scheduling of messages on communication channels or computational tasks on processors presents challenging optimization problem with various constraints and optimization criteria. The talk introduces these combinatorial problems, their mathematical properties, time complexity and algorithms creating a static schedule for time-triggered systems. A general model, based on the Resource Constrained Project Scheduling with Temporal Constraints, is presented first. The approach will be demonstrated on the communication protocols like Profinet IO IRT (an industrial Ethernet protocol standardized in IEC61158) and IEEE 802.15.4/ZigBee (beacon enabled cluster-tree Wireless Sensor Network). Further, we will focus on a mixed-criticality scheduling problem where each task has a criticality given by a positive integer number and the exact processing time of the task is not known. The schedule with different criticality levels is generated off-line, but its on-line execution switches among the criticality levels depending on the actual values of the processing times. Finally we will concentrate on the cyclic scheduling problem used to decide the cluster activation in a wireless sensor network while exploiting polynomial-time complexity of a specific sub-problem.
24. 11. 2015
prof. RNDr. Michal Kozubek, Ph.D., FI MU
From Quantitative Bioimaging to a Model of the Cell
Abstrakt: From a systems biology perspective, the cell is the principal element of information integration. Therefore, understanding the cell in its spatiotemporal context is the key to unraveling many of the still unknown mechanisms of life and disease. This presentation will review image processing aspects relevant to the quantification of cell morphology and dynamics with an emphasis on cell segmentation and tracking. Special attention will be paid to collaborative efforts of the bioimaging community in algorithm benchmarking and providing open access to the acquisition and analysis services, especially across Europe. Finally, the presentation will describe ongoing efforts to integrate all the acquired vast and diverse information to create a credible model of cell morphology and behavior.
1. 12. 2015
Fotios Liarokapis, Ph.D., FI MU
Interactive Augmented Reality Environments
Abstrakt: Augmented reality is considered to be as one of the most promising technologies for the coming years. This talk will present augmented reality technologies for both indoor and outdoor environments. A number of novel augmented reality interfaces will be demonstrated illustrating how these technologies can be used effectively for different application domains such as archaeology, education and navigation.
8. 12. 2015
prof. Rusins Freivalds, Faculty of Physics and Mathematics, University of Latvia, Riga
Ultrametric algorithms and automata
Abstract: Ultrametric algorithms are similar to probabilistic algorithms, only the degree of indeterminism is presented not by a real number between 0 and 1 (called “probability”) but by a p-adic number. This notion was introduced in 2012. Since then many results have been obtained. Partly, they explain why p-adic numbers are so popular in string theory and modern molecular biology. Probably, the most interesting part of this study is the complexity. For some languages ultrametric algorithms are more efficient than probabilistic ones, while for other languages probabilistic algorithms perform better.
15. 12. 2015
RNDr. Nikola Beneš, Ph.D., FI MU
Koncepce výuky programování na FI
Abstrakt: Od absolventů Fakulty informatiky se mimo jiné očekává schopnost programovat, a to v plném rozsahu tohoto termínu. Současná situace tomuto očekávání příliš neodpovídá. Setkáváme se s tím, že zaměstnavatelé hodnotí programátorské schopnosti absolventů FI spíše negativně. Považujeme proto za nutné zamyslet se nad současným stavem výuky programování a připravit návrh úprav, které zvýší dovednosti absolventů v této důležité oblasti.

V první části přednášky se podíváme na současný stav výuky programování na fakultě od všeobecných základů až po výuku konkrétních programovacích jazyků a návazné specializované předměty. V druhé části identifikujeme možná slabá místa současného přístupu a představíme návrh řešení včetně etap postupné realizace. Tato část by měla vyústit v širokou diskusi k tématu výuky programování na FI.