I am part-time professor at the Department of Electronics and Informatics (ETRO) of the Faculty of Applied Sciences (TW) at the Vrije Universiteit Brussel (VUB). My job consists of research and teaching. See Parallel Website for more information about my professional activities.
I am part-time lecturer at the Erasmus University College, Department of Industrial Sciences and Technology.
ResearchFields: Parallel Performance, Causal Performance Models and Causal Structure Learning.
Nowadays less actif: Parallel Processing in general, Parallel Discrete Event Simulation, Parallelism & Intelligence, Complexity & Symmetry.
Stijn Meganck and me founded BN@work in 2010: the European Society for Researchers on Probabilistic Graphical Models.
- 2011 - 2013. The GUDI Project: A combined GP-GPU/FPGA desktop system for accelerating image processing applications.
- Responsibility: Acquire the knowhow about GPUs for lowering the threshold for companies (target: medical sector, image processing).
- 2010 - 2013. The Exascience Project: Solar Flare Prediction on Intel’s Future Exascale Supercomputer.
- Responsibility: Performant visualization of fine-grain massively parallel simulation.
- 2009 - 2013. The use of dynamic models for Prognostics for Optimal Maintenance (POM).
- Responsibility: learning dynamic graphical probabilistic models (such as dynamic Bayesian networks, graphical duration models or hidden Markov models) from data about the degradation process of industrial machines.
- 2008 - 2009. DMOBISA: on Distributed Mosaicing of Medical Image Sequences.
- Responsibility: a transparaent solution for parallelizing the image processing based on clusters, multicores and GPUs.
Journals and book chapters
- Alexander Statnikov, Nikita I. Lytkin, Jan Lemeire, Constantin F. Aliferis, Algorithms for Discovery of Multiple Markov Boundaries, Journal on Machine Learning Research (JMLR), 2013.
- Jan Lemeire, Dominik Janzing, Replacing Causal Faithfulness with Algorithmic Independence of Conditionals, Minds and Machines, 2012, DOI 10.1007/s11023-012-9283-1.
Jan Lemeire, Stijn Meganck, Francesco Cartella, Tingting Liu, Conservative Independence-Based Causal Structure Learning in Absence of Adjacency Faithfulness, International Journal of Approximate Reasoning (IJAR), 2012. Jan Winter, Jan Lemeire, Stijn Megank, Jo Geboers, Gina Rossi, Andreas Mokros, Comparing the Predictive Accuracy of Case Linkage Methods in Serious Sexual Assaults, Journal of Investigative Psychology and Offender Profiling (JIPOP), 2012.
Dominik Janzing, Joris Mooij, Kun Zhang, Jan Lemeire, Jakob Zscheischler, Povilas Daniusis, Bastian Steudel, Bernhard Scholkopf, Information-geometric approach to inferring causal directions, Artificial Intelligence, 2012.
- Check the online java applet showing the experimental results.
Jan Lemeire, Kris Steenhaut, Abdellah Touhafi, When are Graphical Causal Models not Good Models? In Causality in the Sciences, J. Williamson, F. Russo and P. McKay, editors, pages 562-582, Oxford University Press, March 2011.
- Studies causal models and causal inference with the concept of Kolmogorov Minimal Sufficient Statistic.
- Presented at CAPITS 2008. Slides of presentation.
Walter Colitti, Kris Steenhaut, Didier Colle, Mario Pickavet, Jan Lemeire and Ann Nowé, Integrated routing in GMPLS based IP/WDM networks, in Photonic Network Communications, 2010. Jan Lemeire, Kris Steenhaut. Inference of Graphical Causal Models: Representing the Meaningful Information of Probability Distributions, In JMLR Proceedings, Volume 6: Causality: Objectives and Assessment (NIPS 2008), 2010. Abdellah Touhafi, Kris Steenhaut, Jan Lemeire, De watertoren: een integratief project voor toekomstige ingenieurs, in Onderwijsvernieuwing: een continu proces, Thea Derks, Jan Driesen, Arnout Horemans, Frederik Questier, Kris Steenhaut en Hilde Van Lindt (eds.), VubPress, 2008. (Onderwijsvernieuwing & OnderwijsServiceCentrum) Jan Lemeire, Erik Dirkx, Walter Colitti, Modeling the Performance of Communication Schemes on Network Topologies, Parallel Processing Letters, Vol. 18, No. 2, 2008. Jan Lemeire, Erik Dirkx, Frederik Verbist, Causal Analysis for Performance Modeling of Computer Programs. Scientific Programming, Vol. 15, No 3, pp. 121-136, IOS Press, 2007. Jan Lemeire et al., Adaptive Load Balancing of Parallel Applications with Multi-Agent Reinforcement Learning on Heterogeneous Systems, Scientific Programming journal, Vol 12, No 2, 2004.
Conferences and workshops
- Petar Marendic, Jan Lemeire, Tom Haber, Dean Vucinic, Peter Schelkens, An Investigation into the performance of reduction algorithms under load imbalance, in Proceedings of International European Conference on Parallel and Distributed Computing (Euro-Par), Greece, 2012.
- Francesco Cartella, Tingting Liu, Stijn Meganck, Jan Lemeire and Hichem Sahli, Online adaptive learning of Left-Right Continuous HMM for bearings condition assessment, Proceedings of COMADEM 2012.
- Tingting Liu, Jan Lemeire, Francesco Cartella, Stijn Meganck, An improved segmentation-based HMM learning method for Condition-based Maintenance, Proceedings of COMADEM 2012.
- Ahmed Mabrouk, Jan Lemeire, Rune Erlend Jensen, Analyse causale de performance des algorithmes à partir de données d’observation, in Proceedings of 6èmes Journées Francophones sur les Réseaux Bayésiens (JFRB), Nantes, 10-11 Mai 2010.
- Tom Haber, Petar Marendic, Dean Vucinic, Jan Lemeire, Philippe Bekaert. Exascale In-Situ Visualization using Raytracing, International Conference for High Performance Computing, Networking, Storage and Analysis (SC11), Seattle, 2011.
- Jan G. Cornelis, Jan Lemeire. Benchmarks Based on Anti-Parallel Patterns for the Evaluation of GPUs, Parallel Computing Conference (ParCo), Ghent, Belgium, September 2011.
- Jan Lemeire, Stijn Meganck, Francesco Cartella, Tingting Liu and Alexander Statnikov, Inferring the Causal Decomposition under the Presence of Deterministic Relations, Special session Learning of causal relations at the ESANN conference, Bruges, Belgium, April, 2011.
- Jan Lemeire, Stijn Meganck, Francesco Cartella, Robust Independence-Based Causal Structure Learning in Absence of Adjacency Faithfulness, in Procs of European Workshop on Probabilistic Graphical Models (PGM), Helsinki, Finland, 2010.
- Jan Lemeire, Causal structure learning and inductive inference based on Kolmogorov complexity, in Dagstuhl Seminar Proceedings, Machine learning approaches to statistical dependences and causality, September, 2009.
- Jan Lemeire, Yan Zhao, Peter Schelkens, Steve De Backer, Frans Cornelissen and Bert Torfs, Towards Fully User Transparent Task and Data Parallel Image Processing, in Procs. of Workshop on Parallel and Distributed Computing in Image Processing, Video Processing, and Multimedia, ISPA Symposium, 2009.
- Jan Lemeire, Kris Steenhaut, Constraint-based Causal Structure Learning when Faithfulness Fails, Annual machine learning conference of Belgium and The Netherlands (BeneLearn 2009), Tilburg, The Netherlands, 2009.
- Frans Cornelissen, Steve De Backer, Jan Lemeire, Bert Torfs, Rony Nuydens, Theo Meert, Peter Schelkens and Paul Scheunders. Fibered fluorescence microscopy (FFM) of intra epidermal nerve fibers--translational marker for peripheral neuropathies in preclinical research: processing and analysis of the data. In Proc. of Applications of Digital Image Processing XXXI, part of SPIE Symposium on Optical Engineering and Applications, August 2008, San Diego, CA USA.
- Frans Cornelissen et al., Fibered fluorescence microscopy of intra epidermal nerve fibers as translational marker for peripheral neuropathies in preclinical research – Processing and analysis of the data, Knowledge for Growth 2008, Gent, Belgium.
- An Alternative Approach for Playing Complex Games like Chess, Annual machine learning conference of Belgium and The Netherlands (BeneLearn 2008), Spa, Belgium 2008. (slides of Presentation)
- Longer text: An Alternative Approach for Playing Complex Games like Chess: Evaluating The Effects of Patterns by Falsification
Colitti Walter, Steenhaut Kris, Nowe Ann, Lemeire Jan, Multilayer Quality and Grade of Service Support for High Speed GMPLS IP/DWDM Networks, NBiS 2007, LNCS, Volume: 4658, pp: 187 - 196, 2007.
The Representation and Learning of Equivalent Information in Causal Models. Technical Report IRIS-TR-0099, May 2006.
- Proposes a solution to the well-known problem that deterministic relations cannot be represented by faithful Bayesian Nets.
Causal Performance Models of Computer Systems: Definition and Learning Algorithms. Technical Report IRIS-TR-0100, 2006.
- About the utilization of causal models in performance analysis.
A Refinement Strategy for a User-Oriented Performance Analysis. (Euro-Pvm 2004, slides of presentation) Causal Models for Parallel Performance Analysis, Fourth PA3CT-Symposium, Edegem, Belgium, September 2004.
Lookahead Accumulation in Conservative Parallel Discrete Event Simulation. The 2004 High Performance Computing & Simulation (HPC&S) Conference.
Exploiting Symmetrical Properties for Partitioning of Models in Parallel Discrete Event Simulation. (PADS 2004, Kufstein, Austria)
Complexity-Preserving Functions (DIMACS Workshop on Complexity and Inference 2003, presentation 'Complexity and Symmetry' ) Automated Experimental Parallel Performance Analysis (2002) (2nd PACT Symposium 2002) Adaptive Load Balancing of Parallel Applications with Reinforcement Learning on Heterogenous Networks (2002) (DCABES 2002) Performance Factors in PDES (2001) (ESM conference 2001)Unpublished
- Causal Inference on Data Containing Deterministic Relations, February 2008.
Causal Models as Minimal Descriptions of Multivariate Systems, 2006.
- The results of this initial work has, finally, been published: see our Mind&Machines paper.
Talk 'Practical Parallel Processing' at Royal Military Academy, May 2004, Brussels.
Causes of Blocking Overhead in Message-Passing Programs (2003) Towards a Generalised Performance Analysis of Parallel Processing (2003) Natuurlijke Taal in de Formele Wereld van de Informatica (2001) Neural Networks: What's Inside. The Explicitness Hypothesis (2001)
PhDThe work presented in my thesis consists of a philosophical, theoretical and practical exploration of causal inference and its benefits for the performance analysis of (parallel) computer programs.
PhD final text promotion text (ook in het Nederlands) and abstract
Presentation given at the public defense (19 december 2007)
Listen to the recorded talk: Evaluation of causal discovery with Bayesian networks with the principle of Kolmogorov Minimal Sufficient Statistic, Thursday 16 October 2008
Website of machine learning reading group
Introduction to Bayesian networks, presentation given at Verhaert, 30th january 2009 (Up2Date seminars)
Seminar at Machine Learning Group of the ULB, 15th November 2006 (MLG-ULB)General talk about my PhD research, 24th May, 2006 (one of the weekly ETRO seminars)
I am responsible for the following courses:
I'm the mentor of several students doing projects or their final thesis (see students section of the lab's website).
- 1st year bachelors course 'Informatica': Python (semester 1) and Java, Algorithms and Data structures (semester 2).
- 3rd year bachelors: Operating Systems
- 3rd year bachelors: Advanced Programming Techniques
- Parallel Systems
- Advanced Computer Architecture
- Computer Architecture
- Digital Signal Processing.
Vrije Universiteit Brussel (VUB-map, find the VUB)
Faculty of Applied Sciences, ETRO dept.
Pleinlaan 2, B-1050 Brussels, Belgium
I'm occupying 2 offices, best is to make an appointment and make sure that it's clear where we will meet
Parallel computing: Pleinlaan 9, second floor, room PL9-28 (how to find me and pleinlaan 9)
Probabilistic graphical models: Building Ke, room Ke3.08 (entrance via building K, third floor, follow the arrows)
Tel +32 2 629.16.79
Fax +32 2 629.28.70
Email : firstname.lastname@example.org
I graduated from the VUB in 1994 and received my (masters) diploma of electrotechnics engineer. I did my thesis at the VUB Artificial Intelligence-lab in the context of expert systems. Then I completed my studies with an additional masters degree in Computer Science (Faculty of Science), also at the VUB, in 1995.
At the start of my professional carreer, I worked for 3.5 years in the private sector: first as a programmer for Cap Gemini, an IT consulting firm, then for Warmoes & Van Damme, a company specialised in knowledge systems (now, partly Aktor). I developed my professional skills during these years, but also found out that research is my real passion. Therefore I returned to the academic world to prove myself in a scientific carreer...
At first I was allocated on a project with Alcatel Bell on parallel simulation. Since march 2001 I was employed as an assistent, teaching with a lot of enthusiasm and combining it with pursuing a PhD about parallel performance analysis and causal inference. In December 2007 I received my PhD and got a postdoc position as VUB doctor-assistant. Since then I participate in the IBBT-project DMOBISA, focussing on the parallelization of image processing algorithms.
Besides my job, my life consists mainly of: