Michal Gregor
Researcher – Expert
Kempelen Institute of Intelligent Technologies (KInIT)
University of Žilina (UNIZA)
Biography
I am a researcher at the Kempelen Institute of Intelligent Technologies and a researcher at the Institute of Competitiveness of the University of Žilina – focusing, in both capacities, on the area of machine learning. In addition, I am currently coordinating the Laboratory of Artificial Intelligence of the University of Žilina (LUIZA).
- Artificial Intelligence
- Machine Learning
- Deep Learning
- Multi-modal Learning
- Reinforcement Learning
- Data Analysis
- …
-
PhD. in Automation, 2014
University of Žilina
-
Master's Degree (Ing.) in Automation, 2011
University of Žilina
-
Bachelor's Degree (Bc.) in Automation, 2009
University of Žilina
Skills
pytorch, scikit-learn, pandas, seaborn, numpy, pyro, …
deep learning; reinforcement learning; metaheuristics; decision trees; …
Python, C++, C, PHP, Matlab, JavaScript, …
LaTeX, HTML, CSS, XML, Markdown, …
git, svn, hg, …
Windows, Linux
Experience
Activities:
- Language models and multi-modal learning;
- DisAI: Improving scientific excellence and creativity in combating disinformation with artificial intelligence and language technologies. HORIZON-WIDERA-2021-ACCESS-03-01 - Twinning: 101079164;
- Commercial projects;
- …
Activities:
- KEGA 008ŽU-4/2021: Integrated Teaching for Artificial Intelligence Methods at the University of Žilina.
- Innovative autonomous logistic vehicle, NFP313010P904.
- Research in the area of machine learning and artificial intelligence.
- VEGA 1/0938/16: An adaptive system of internal logistics (ASIL).
- …
Activities:
- Development project 002ZU-2-1-2021: Hybrid Education in the area of artificial intelligence, machine learning and cybernetics at UNIZA;
- …
A 6-month research stay at the UCB. Participation in VEHICAL:
- BAIR (Berkeley Artificial Intelligence Research) laboratory;
- HART (Human-Assistive Robotic Technologies) laboratory.
Human driver modeling, psychoacoustic aspects of driving, deep reinforcement learning for decreasing psychoacoustic annoyance.
Activities:
- KEGA 014ŽU-4 / 2018: Broadening the content in a field of study with respect to the current requirements of the industry as regards artificial intelligence methods and IT (deputy project leader).
- H2020-MSCA-RISE-2016, 734331: SENSIBLE – SENSors and Intelligence in BuiLt Environment.
- KEGA 038ŽU-4 / 2017: Laboratory methods of teaching automatic identification and localization using radio-frequency identification technology.
- APVV-16-0006: Automated robotic assembly cell as an instrument of concept Industry 4.0 (project coordinator for the University of Žilina).
- KEGA 010ŽU-4 / 2013: Modernization of didactic equipment and teaching methods with a focus on the area of robotics.
- VEGA 1 / 0453 / 12 Study of interactions of a motor vehicle, traffic flow and road.
- …
Activities:
- The University Science Park of the University of Žilina. ITMS 26220220184.
Activities:
- Development of a voice control system.
- VMROS – Development of modular mobile robotic systems, investigator of activity 2.1 “Design and optimization of modular robotic configurations in virtual reality”.
- ZATIPS – Competence Center for Knowledge technology innovation of production systems in industry and services. ITMS: 26220220155.
Activities:
- KEGA 010ŽU-4 / 2013: Modernization of didactic equipment and teaching methods with a focus on the area of robotics.
- VEGA 1 / 0453 / 12 Study of interactions of a motor vehicle, traffic flow and road.
- Centre of excellence for systems and services of intelligent transport. University of Žilina. ITMS 26220120028, ITMS 26220120050.
- New methods of measuring physical dynamic parameters and interactions of motor vehicles, in cooperation with Betamont, s.r.o. ITMS: 26220220089.
- Activity 1.3 Centre of telematic systems at the University of Žilina, project Research Center of Traffic Telematics. Coordinator: BETAMONT s.r.o. ITMS: 26220220169.
Recent & Upcoming Talks
Teaching
lectures, seminars and labs
University of Žilina
- Machine Learning – lectures, seminars (Summer ‘2023).
- Artificial Intelligence 1 – labs and selected lectures (Summer ‘2016 – ‘2022).
- Computer Vision – lectures, labs, projects (Winter ‘2022); lectures + projects (Winter ‘2023).
- Fundaments of Data Analysis – lectures, seminars (Winter ‘2022).
- Artificial Intelligence – labs (Winter ‘2022).
- Artificial Intelligence 2 – labs (Winter ‘2016, ‘2017).
- Artificial Intelligence – seminars and labs (Summer ‘2011 – ‘2015).
- Expert Systems – labs (Winter ‘2011 – ‘2015).
Erasmus
- Intro to AI, deep learning, embeddings and face clustering (Summer ‘2019).
- Deep learning, reinforcement learning (Summer ‘2015, ‘2017).
Textbooks
- Gregor, M. Umelá inteligencia 1 [Artificial Intelligence 1]. Žilina: CEIT, a.s., 2014. ISBN 978-80-971684-1-4;
- Gregor, M. – Nemec, D. – Hruboš, M. – Spalek, J. Umelá inteligencia 2: Hlboké učenie [Artificial Intelligence 2]. CEIT, a.s., 2017. ISBN 978-80-89865-03-1;
- Gregor, M. – Hruboš, M. – Nemec, D. Umelá inteligencia, skriptá I: Návody na vybrané cvičenia [Artificial Intelligence, Lecture Notes I]. CEIT, a.s., 2017. ISBN 978-80-89865-02-4;
- Gregor, M. – Janota, A. – Hruboš, M. Kompendium vybraných metód umelej inteligencie. EDIS: Vydavateľstvo Žilinskej univerzity, 2019.
Students
thesis supervisor, interns, mentoring, …
Interns
@KInIT
-
Pavol Kebis (2024):
Graduate of the Commenius University in Bratislava and University of Oxford, interning at KInIT; Topic: Efficient adaptation of large language models to low-resource languages. -
Kristína Sásiková (2024):
Student of the Commenius University in Bratislava participating in the DisAI Replication Challenge; Topic: Replicating paper Scaling Sentence Embeddings with Large Language Models. -
Michael Pavlík (2023):
Intern from Tilburg University; Topics: unsupervised reasoning and self-correction in LLMs.
@UNIZA
-
Gaétan Becker (2022):
Intern from Télécom SudParis. Topics: Deep learning architectures for predicting a quantity from other data during sensor outage. -
Sylvain Courty (2019):
Intern from Université d’Angers. Topics: Probabilistic generation of synthetic data for visual object detection. -
Daniel Adamkovič (2017–2018):
Intern from UNIZA. Topics: Machine learning, deep learning, reinforcement learning, …
More...
Undergrads @UNIZA
-
Erik Dobeš (2021 / 2022):
BSc. thesis: Controlling a balancing robot using artificial intelligence methods (In Slovak: Riadenie balansujúceho robota s podporou metód umelej inteligencie). -
Andrej Kováč (2018 / 2019):
BSc. thesis: Scraping Information from the Web and Summarizing it in a Formatted Output (In Slovak: Scraping informácií z web-u a ich sumarizácia vo formátovanom výstupe). -
Martin Vít (2017 / 2018):
BSc. thesis: Implementation of Artificial Players for 2048 (In Slovak: Implementácia umelých hráčov pre hru 2048). -
Lukáš Slováček (2015 / 2016):
BSc. thesis: Chatbot Systems (In Slovak: Chatbot systémy).
Graduates @UNIZA
-
Filip Kalus (2022 / 2023):
MSc. thesis: Traffic Signal Control using Deep Reinforcement Learning Methods (In Slovak: Riadenie dopravy pomocou metód hlbokého učenia s odmenou). -
Martin Vít (2019 / 2020):
MSc. thesis: An Interface for Speech Recognition Using Deep Learning (In Slovak: Rozhranie pre rozpoznávanie reči pomocou metód hlbokého učenia). -
Martin Vojtyla (2018 / 2019):
MSc. thesis: Solving Scheduling Problems using MiniZinc (In Slovak: Riešenie rozvrhovacích problémov pomocou nástroja MiniZinc). -
Lukáš Slováček (2017 / 2018):
MSc. thesis: Simulation of Traffic and its Control using SUMO (In Slovak: Simulácia dopravy a jej riadenia pomocou nástroja SUMO). -
Matej Lokaj (2017 / 2018):
MSc. thesis: Grammar Checking based on Recurrent Neural Networks (In Slovak: Kontrola pravopisu na báze rekurentných neurónových sietí). -
Jakub Hanes (2016 / 2017):
MSc. thesis: Reinforcement Learning (In Slovak: Učenie s odmenou). -
Juraj Kindernay (2016 / 2017):
MSc. thesis: Rebuilding of the graphical user interface for the Fuzzylite library (In Slovak: Prestavba grafického užívateľského rozhrania pre knižnicu Fuzzylite). -
Dávid Bujňák (2015 / 2016):
MSc. thesis: Implementation and Solution of a Planning Problem (In Slovak: Implementácia a riešenie plánovacieho problému). -
Michal Jombík (2013 / 2014):
MSc. thesis: Chatbot Systems (In Slovak: Chatbot systémy). -
Michal Kubica (2017 / 2018):
MSc. thesis: A Robotic Workplace for Washing Machine Assembly (In Slovak: Robotizované pracovisko montáže automatickej práčky). Co-advisor: Ing. Anton Tomáš. -
Vladimír Cingel (2015 / 2016):
MSc. thesis: Visualization and Control of a Virtual Industrial Robot (In Slovak: Vizualizácia a riadenie virtuálneho priemyselného robota). Co-advisor: Ing. Tomáš Michulek, PhD. -
Lukáš Hrčka (2015 / 2016):
MSc. thesis: A Monitoring System for an Automated Industrial Logistic System (In Slovak: Monitorovací systém autonómnej podnikovej logistiky). Co-advisor: Ing. Tomáš Michulek, PhD. -
Miloš Kapišovský (2015 / 2016):
MSc. thesis: Control of an Industrial Robot using a PLC Controller (In Slovak: Riadenie priemyselného robota prostredníctvom PLC automatu). Co-advisor: Ing. Peter Marčan.
Service
service to the community, committees, events, memberships, …
Memberships
- 2023: Honorary member of the National Centre of Robotics;
- 2023: Honorary member of the Department of Control and Information Systems, University of Žilina;
- 2021: Member of the executive board of association AI4SK;
Service
- 2019: Founding of the Laboratory of Artificial Intelligence of the University of Žilina (LUIZA);
- 2019: Cooperation on the preparation of the Slovak national strategy for artificial intelligence with Slovak.AI
- 2019—… Co-organizing the annual Machine Learning Summer School at the University of Žilina;
- 2014—… Reviewer: Applied Soft Computing, Computing and Informatics, ICANN, ELEKTRO, TRANSCOM, C&I, …
Theses
my qualification theses
The thesis proposes multiple approaches that improve the speed of learning in the context of reinforcement learning. It shows multiple ways in which prior knowledge can be utilized. It considers both external objectives and intrinsic motivational signals in a single multi-objective motivational system.
The thesis deals with the application of artificial intelligence methods – especially of genetic algorithms and genetic programming – to design and control of robotic systems. It proposes several adaptive mechanisms and shows in what ways these improve upon vanilla genetic programming.
The thesis deals with automatic speech recognition systems. It develops and verifies a system for recognition of isolated speech units, based on artificial neural networks.
Publicity
-
Denník N: Ocenili priekopníkov umelej inteligencie. Nobelovu cenu udelili za fyziku neurónových sietí.
-
TREND: Jazykové modely zatiaľ nedokážu spoľahlivo nahradiť ľudí. V praxi ale do toho firmy možno pôjdu.
-
SHARE podcast, zive: Expert na AI: Pri odhaľovaní deepfake pomôžu nástroje len z časti. Najdôležitejšie bude kriticky myslieť (podcast).
-
Aktuality: Kompro na osobnosti, nahé celebrity: Deepfake sa výrazne zlepšil, prichádza éra neistoty.
-
Rádio_FM Generovanie fotografií.
-
Denník N: Úzkosti, depresia či strata sebaúcty. Čo môžu spôsobiť dokonalé skrášľovacie filtre?.
-
zive: Ako zmení umelá inteligencia slovenské školstvo? Pýtali sme sa odborníkov.
-
zive: Letná škola strojového učenia vyvrcholila odovzdávaním certifikátov.
-
teraz.sk Letná škola strojového učenia UNIZA vzdeláva prezenčne aj online.
-
SME.sk: Letná škola strojového učenia – nová forma vzdelávania.
-
TA3: Slovensko má nedostatok študentov IT, odborníci chcú prilákať ďalších (TA3).
-
Týždeň: Aby sme sa neocitli v dystopii.
-
Spravodajca: Od Terry Pratchetta k programovaniu.
Contact
- michal@gregor.sk
- Univerzitná 1, Zilina, 01322; Slovak Republic
- Connect on LinkedIn
- michal.gregor@kinit.sk
- michal.gregor@uniza.sk