
Teaching Activity



2019 – 
Mathematical Foundations of Machine Learning (MFML), Audience: M.Sc. and Ph.D. Students in Applied Mathematics, Institute of Mathematics, Eötvös Loránd University (ELTE), Budapest, Hungary
[syllabus] [slides]



2019 – 
Markov Decision Processes and Reinforcement Learning (MDPs & RL), Audience: M.Sc. and Ph.D. Students in Mathematics, Institute of Mathematics, Budapest University of Technology and Economics (BME), Budapest, Hungary
[slides]



2015 – 
Stochastic Models and Adaptive Algorithms, Ph.D. School of Computer Science, Eötvös Loránd University (ELTE), Budapest, Hungary, and (since 2017) Budapest University of Technology and Economics (BME), Budapest, Hungary
[slides]



2013 – 2014 
Mathematical Optimization (main organizer), SZTAKI: Institute for Computer Science and Control, Budapest, Hungary



2012 
Probability and Random Models (ELEN90054, with Girish Nair), School of Engineering, University of Melbourne, Australia



2005 – 2006 
Markov Decision Processes (organized by Cs. Szepesvári), SZTAKI: Institute for Computer Science and Control, Hungary



2002 
Theory of Operating Systems, Department of Information Systems, Eötvös Loránd University (ELTE), Budapest, Hungary



2000 – 2002 
Programming Methodology, Department of Software Technology, Eötvös Loránd University (ELTE), Budapest, Hungary



Lecture Slides



2017 
On the Reliability of Regression Models, Publication Award Seminar, SZTAKI, Budapest, Hungary [slides]



2012 
DistributionFree System Identification: Exact, NonAsymptotic Confidence Regions, Faculty of Electrical Engineering, Computer Science, and Mathematics, University of Paderborn, Germany [slides]



2010 
Introduction to Markov Decision Processes, Department of Electrical Engineering, University of Melbourne, Australia [slides]



2009 
A Machine Learning Approach to Stochastic Resource Control, Poster, DYSCO Study Day, Mons, Belgium [poster]



2008 
Learning in Changing Environments: Reinforcement Learning in Environments with Asymptotically Bounded Variation, Gatsby Computational and Theoretical Neuroscience and Machine Learning Unit, University College London (UCL), UK
[slides]



2006 
Introduction to OffPolicy Learning, MDP Seminar, SZTAKI, Budapest, Hungary [slides]



2005 
Introduction to Temporal Difference Learning (Hungarian Slides), MDP Seminar, SZTAKI, Budapest, Hungary [slides]



2004 
Intuitionism in Mathematics (Philosophy of Mathematics), HalSzem, ELTE, Budapest, Hungary [pdf] Handout [pdf]




Selected Journal Papers



– 
Csáji, B. Cs.; Kis, K. B.: DistributionFree Uncertainty Quantification for Kernel Methods by Gradient Perturbations, Machine Learning, Springer, Special Issue of the European Conference on Machine Learning (ECML PKDD Journal Track), Vol. 108, 2019, pp. 1677–1699
[pdf]
[link]



– 
Weyer, E.; Campi, M. C.; Csáji, B. Cs.: Asymptotic Properties of SPS Confidence Regions, Automatica, Elsevier, Vol. 82, 2017, pp. 287–294
[pdf]



– 
Csáji, B. Cs.; Kemény, Zs.; Pedone, G.; Kuti, A.; Váncza, J.: Wireless MultiSensor Networks for Smart Cities: A Prototype System with Statistical Data Analysis, IEEE Sensors Journal, IEEE Press, Vol. 17, Issue 23, 2017, pp. 7667–7676 [arxiv]



– 
Carè, A.; Csáji, B. Cs.; Campi, M. C.; Weyer, E.: FiniteSample System Identification: An Overview and a New Correlation Method, IEEE Control Systems Letters (LCSS), IEEE Press, Vol. 2, No. 1, 2017, pp. 61–66 [pdf]



– 
Kovács, A.; Bátai, R.; Csáji, B. Cs.; Dudás, P.; Háy, B.; Pedone, G.; Révész, T.; Váncza, J.: Intelligent Control for EnergyPositive Street Lighting, Energy: The International Journal,
Elsevier, Vol. 114, 2016, pp. 40–51 [pdf]



– 
Csáji, B. Cs.; Campi, M. C.; Weyer, E.: SignPerturbed Sums: A New System Identification Approach for Constructing Exact NonAsymptotic Confidence Regions in Linear Regression Models, IEEE Transactions on Signal Processing, IEEE Press, Vol. 69, 2015, pp. 169–181 [pdf]



– 
Monostori, L.; Valckenaers, P.; Dolgui, A.; Panetto, H.; Brdys, M.; Csáji, B. Cs.: Cooperative Control in Production and Logistics, Annual Reviews in Control (ARC): A Journal of IFAC, the International Federation of Automatic Control, Elsevier, Vol. 39, 2015, pp. 12–29 [pdf]



– 
Csáji, B. Cs.; Jungers, R. M.; Blondel, V. D.: PageRank Optimization by Edge Selection, Discrete Applied Mathematics (DAM): The Journal of Combinatorial Algorithms, Informatics and Computational Sciences, Elsevier, Vol. 169, 2014, pp. 73–87 [pdf]



– 
Csáji, B. Cs.; Browet, A.; Traag, V. A.; Delvenne, JC.; Huens, E.; Van Dooren, P.; Smoreda, Z.; Blondel, V. D.: Exploring the Mobility of Mobile Phone Users, Physica A: Statistical Mechanics and its Applications, Elsevier, Vol. 392, Issue 6, 2013, pp. 1459–1473 [pdf]



– 
Csáji, B. Cs.; Monostori, L.: Adaptive Stochastic Resource Control: A Machine Learning Approach, Journal of Artificial Intelligence Research (JAIR), AAAI Press, Vol. 32, 2008, pp. 453–486 [pdf]



– 
Csáji, B. Cs.; Monostori, L.: Value Function Based Reinforcement Learning in Changing Markovian Environments, Journal of Machine Learning Research (JMLR), MIT Press and Microtome Publishing, Vol. 9, 2008, pp. 1679–1709 [pdf]



Selected Conference Papers



– 
Csáji, B. Cs.; Kis, K. B.; Kovács, A.: A SamplingandDiscarding Approach to Stochastic Model Predictive Control for Renewable Energy Systems, 21st IFAC World Congress (1st Virtual IFAC World Congress), accepted / in press, July 11–17, 2020 [pdf] [slides]



– 
Csáji, B. Cs.; Tamás, A.: SemiParametric Uncertainty Bounds for Binary Classification, 58th IEEE Conference on Decision and Control (CDC), Nice, France, December 11–13, 2019, pp. 4427–4432 [pdf] [slides]



– 
Carè, A.; Csáji, B. Cs.; Gerencsér, B.; Gerencsér, L.; Rásonyi, M.: ParameterDependent Poisson Equations: Tools for Stochastic Approximation in a Markovian Framework, 58th IEEE Conference on Decision and Control (CDC), Nice, France, December 11–13, 2019, pp. 2259–2264 [pdf]



– 
Csáji, B. Cs.; Kis, K. B.: DistributionFree Uncertainty Quantification for Kernel Methods by Gradient Perturbations, 58th European Conference on Machine Learning (ECML PKDD), Würzburg, Germany, September 16–20, 2019 [arxiv] [slides] [poster]



– 
Gerencsér, L.; Csáji, B. Cs.; Sabanis, S.: Asymptotic Analysis of the LMS Algorithm with Momentum, 57th IEEE Conference on Decision and Control (CDC), Miami Beach, Florida, December 17–19, 2018, pp. 3062–3067 [pdf] [slides]



– 
Csáji, B. Cs.: NonAsymptotic Confidence Regions for Regularized Linear Regression Estimates, 20th European Conference on Mathematics for Industry (ECMI), FiniteSample System Identification Minisymposium, Budapest, Hungary, June 18–22, 2018, Springer, pp. 605–611
[pdf]



– 
Kolumbán, S.; Csáji, B. Cs.: Towards DOptimal Input Design for FiniteSample System Identification, 18th IFAC Symposium on System Identification (SYSID), Stockholm, Sweden, July 9–11, 2018, pp. 215–220
[pdf]



– 
Carè, A.; Csáji, B. Cs.; Campi, M. C.; Weyer, E.: FiniteSample System Identification: An Overview and a New Correlation Method, 56th IEEE Conference on Decision and Control (CDC), Melbourne, Australia, December 12–15, 2017, pp. 4612–4617 [pdf] [slides]



– 
Carè, A.; Csáji, B. Cs.; Campi, M. C.; Weyer, E.: Undermodelling Detection with SignPerturbed Sums, 20th IFAC World Congress, Toulouse, France, July 9–14, 2017, pp. 2799–2804 [pdf] [slides]



– 
Carè, A.; Csáji, B. Cs.; Campi, M. C.: SignPerturbed Sums (SPS) with Asymmetric Noise: Robustness Analysis and Robustification Techniques, 55th IEEE Conference on Decision and Control (CDC), Las Vegas, Nevada, December 12–14, 2016, pp. 262–267 [pdf]



– 
Csáji, B. Cs.: Score Permutation Based Finite Sample Inference for Generalized AutoRegressive Conditional Heteroskedasticity (GARCH) Models, 19th International Conference on Artificial Intelligence and Statistics (AISTATS), Cadiz, Spain, May 9–11, 2016, pp. 296–304 [pdf] [poster]



– 
Volpe, V.; Csáji, B. Cs.; Carè, A.; Weyer, E.; Campi, M. C.: SignPerturbed Sums (SPS) with Instrumental Variables for the Identification of ARX Systems, 54th IEEE Conference on Decision and Control (CDC), Osaka, Japan, December 15–18, 2015, pp. 2115–2120 [arxiv]



– 
Csáji, B. Cs.; Weyer, E.: ClosedLoop Applicability of the SignPerturbed Sums Method, 54th IEEE Conference on Decision and Control (CDC), Osaka, Japan, December 15–18, 2015, pp. 1441–1446 [pdf] [slides]



– 
Csáji, B. Cs.; Kovács, A.: Adaptive Aggregated Predictions for Renewable Energy Systems, 2014 IEEE SSCI Symposium on Adaptive Dynamic Programming and Reinforcement Learning (ADPRL), Orlando, Florida, December 9–12, 2014, pp. 132–139 [pdf]



– 
Csáji, B. Cs.; Campi, M. C.; Weyer, E.: Strong Consistency of the SignPerturbed Sums Method, 53rd IEEE Conference on Decision and Control (CDC), Los Angeles, California, December 15–17, 2014, pp. 3352–3357
[pdf]



– 
Weyer, E.; Csáji, B. Cs.; Campi, M. C.: Guaranteed NonAsymptotic Confidence Ellipsoids for FIR Systems, 52nd IEEE Conference on Decision and Control (CDC), Florence, Italy, December 10–13, 2013, pp. 7162–7167
[pdf]



– 
Csáji, B. Cs.; Campi, M. C.; Weyer, E.: SignPerturbed Sums (SPS): A Method for Constructing Exact FiniteSample Confidence Regions for General Linear Systems, 51st IEEE Conference on Decision and Control (CDC), Maui, Hawaii, December 10–13, 2012, pp. 7321–7326 [pdf]



– 
Campi, M. C.; Csáji, B. Cs.; Garatti, S.; Weyer, E.: Certified System Identification: Towards DistributionFree Results, 16th IFAC Symposium on System Identification (SYSID), Brussels, Belgium, July 11–13, 2012, pp. 245–255 [pdf]



– 
Csáji, B. Cs.; Campi, M. C.; Weyer, E.: NonAsymptotic Confidence Regions for the LeastSquares Estimate, 16th IFAC Symposium on System Identification (SYSID), Brussels, Belgium, July 11–13, 2012, pp. 227–232 [pdf]



– 
Csáji, B. Cs.; Weyer, E.: Recursive Estimation of ARX Systems Using Binary Sensors with Adjustable Thresholds, 16th IFAC Symposium on System Identification (SYSID), Brussels, Belgium, July 11–13, 2012, pp. 1185–1190 [pdf] [slides]



– 
Csáji, B. Cs.; Weyer, E.: System Identification with Binary Observations by Stochastic Approximation and Active Learning, 50th IEEE Conference on Decision and Control (CDC) & European Control Conference (ECC), Orlando, Florida, December 12–15, 2011, pp. 3634–3639 [pdf]



– 
Ivanov, T.; Csáji, B. Cs.: Reproducing Kernels Preserving Algebraic Structure: A Duality Approach, 19th International Symposium on Mathematical Theory of Networks and Systems (MTMS), Budapest, Hungary, July 5–9, 2010, pp. 1161–1167 [pdf]



– 
Csáji, B. Cs.; Jungers, R. M.; Blondel, V. D.: PageRank Optimization in Polynomial Time by Stochastic Shortest Path Reformulation, 21st International Conference on Algorithmic Learning Theory (ALT), Canberra, Australia, October 6–8, 2010, pp. 89–103 [pdf]



– 
Csáji, B. Cs.; Monostori, L.: Adaptive Sampling Based LargeScale Stochastic Resource Control, 21st National Conference on Artificial Intelligence (AAAI), Boston, Massachusetts, July 16–20, 2006, pp. 815–820
[pdf]



Selected Papers in Hungarian



– 
Csáji, B. Cs.: Szimmetria és konfidencia (Symmetry and Confidence), Alkalmazott Matematikai Lapok (Applied Mathematical Journal of the Section of Mathematics, Hungarian Academy of Sciences), Vol. 36, No. 2, 2019, pp. 271–278 [pdf]



– 
Csáji, B. Cs.; Rédei, M.: A racionális demokratikus véleményösszegzés korlátairól (On the Limits of Rational Democratic Judgment Aggregation), Magyar Filozófiai Szemle (Hungarian Philosophical Review), Vol. 55, No. 2, 2011, pp. 97–121
[pdf]


