Lukáš Bajer

Lukáš Bajer

doctoral/PhD student

Academy of Sciences of the Czech Republic (AV ČR)
Institute of Computer Science
Department of Machine learning

and

Charles University in Prague, Czech Republic
Faculty of Mathematics and Physics
Department of Theoretical Computer Science

e-mail: bajeluk [aT] seznam [doT] cz

Main research interest: Black-box optimization, Machine learning, Evolutionary algorithms, Surrogate modelling, Statistical models, Gaussian processes

Supervisor: Martin Holeňa
Institute of Computer Science
Academy of Sciences of the Czech Republic


Software

Surrogate CMA-ES, DTS-CMA-ES
The CMA-ES optimization alogorithm by N. Hansen with a Gaussian-process- (and partly also random-forest-) surrogate model written in Matlab.
Repository: github.com/bajeluk/surrogate-cmaes
Authors: Lukáš Bajer (main developer), Zbyněk Pitra, Jakub Repický
License: not decided yet, so all rights for reuse reserved, please contact the author if interested to reuse for other than research/academical purpuses.


Publications

ORCID ScopusID ResearcherID

Bajer, L. and Holeňa, M. Surrogate Model for Continuous and Discrete Genetic Optimization Based on RBF Networks. In Intelligent Data Engineering and Automated Learning – IDEAL 2010, LNCS 6283, p. 251–258, Lecture Notes in Computer Science, Springer, 2010. DOI: 10.1007/978-3-642-15381-5_31

Bajer, L. and Holeňa, M. Surrogate Model for Mixed-Variables Evolutionary Optimization Based on GLM and RBF Networks. In Proceedings of SOFSEM 2013, LNCS 7741, Springer, 2013. DOI: 10.1007/978-3-642-35843-2_41 PDF

Bajer, L. and Holeňa., M. Model Guided Sampling Optimization for Low-dimensional Problems. In Proceedings of ICAART 2015. SCITEPRESS, 2015. Available on arXive.org.

Bajer, L., Pitra, Z. and Holeňa, M. Benchmarking Gaussian Processes and Random Forests Surrogate Models on the BBOB Noiseless Testbed. In Proceedings of the GECCO '15 Companion, ACM, 2015. DOI: 10.1145/2739482.2768468 preprint ACM PDF

Holeňa, M, Bajer, L. and Sčavnický, M. Using Copulas in Data Mining Based on the Observational Calculus. In IEEE Transactions on Knowledge and Data Engineering, 27(10), 2015. DOI: 10.1109/TKDE.2015.2426705

Pitra, Z., Bajer, L. and Holeňa, M. Doubly trained Evolution Control for the Surrogate CMA-ES. In Proceedings of PPSN XIV, LNCS 9921, Springer, 2016. DOI: 10.1007/978-3-319-45823-6_6 PDF

Pitra, Z., Bajer, L., Repický, J. and Holeňa, M. Comparison of Ordinal and Metric Gaussian Process Regression As Surrogate Models for CMA Evolution Strategy. In Proceedings of the GECCO ’17 Companion, ACM, 2017. DOI: 10.1145/3067695.3084206 preprint ACM PDF

Pitra, Z., Bajer, L., Repický, J. and Holeňa, M. Overview of Surrogate-model Versions of Covariance Matrix Adaptation Evolution Strategy. In Proceedings of the GECCO ’17 Companion, ACM, 2017. DOI: 10.1145/3067695.3082539 preprint ACM PDF


Competitions and Awards

Black Box Optimization Competition at GECCO 2017: 1st and 2nd prize with DTS-CMA-ES algorithm


Links

Personal Wiki
bajewiki (mostly in Czech)

Image Gallery
http://picasaweb.google.com/bajeluk

Blog from Finlad (1-year study exchage programme 2006/2007)
Z jižního severu (mostly in Czech)

Personal bookmarks
bookmarks exported from Firefox (mostly in Czech)

My boy-scout group in Liberec (in Czech) http://ichthys.skauting.cz/


Previous studies

Master's degree

Oct 2007 - Sep 2009
Theoretical Computer Science at MFF UK
title: Mgr. (magister, MSc. equivalent)
Diploma thesis: Improving evolutionary algorithms using probability mixture models (in Czech) [PDF]

Related works:

Bachelor's degree

Oct 2003 - Sep 2007
Computer Science (programming) at MFF UK
title: Bc. (bachelor)
Bachelor's thesis: Movement in project ENTs (in Czech) [PDF]


last updated: 15.11.2017
Valid XHTML 1.0 Transitional