Dr. Saso Dzeroski

Jozef Stefan Institute
Department of Intelligent Systems
Jamova 39, 1111 Ljubljana, Slovenia
Phone: (+386)(61) 177 3217, Fax: (+386)(61) 1251 038
E-mail: Saso.Dzeroski@ijs.si

Hot stuff

• Analysis of environmental data with machine learning methods
• Inductive logic programming
• Knowledge discovery in databases (in Slovenian)
• Inductive logic programming and knowledge discovery in databases
• Sixteenth International Conference on Machine Learning

The usual stuff

• Research
• Curriculum vitae
• Events
• Machine learning systems
• Memberships
• Projects
• Publications

Research

My research interests are mainly in the field of machine learning.
Some of them (listed in alphabetical order) are:

• Applications of machine learning
• Computational learning theory
• Inductive logic programming
• Machine discovery
• Knowledge discovery in databases

A summary of topics addressed in my research

Theory:

• Formalization of learning clausal theories from interpretations
• PAC-learnability of logic programs
• PAC-learnability of clausal theories

Methods:

• Discovering differential equations from measured data
• Handling noise in ILP
• Handling real numbers in ILP

Practical applications in the domains of:

• Ecology
  • Analysis of water quality data on Slovenian rivers
  • Biological classification of British rivers
  • Modelling algal growth in the Lagoon of Venice
• Medicine
  • Early diagnosis of rheumatic diseases
  • Treating acute abdominal pain in children
• Pharmacology
  • Diterpene structure elucidation from NMR spectra

Curriculum vitae

Born 31st May 1968, Ohrid , Macedonia

• 1985. Graduated from high school.
• 1989. BSc in Computer Science.
  Faculty of Electrical Engineering and Computer Science, University of Ljubljana, Slovenia.
  Thesis: Controlling the inverted pendulum.
• 1991. MSc in Computer Science.
  Faculty of Electrical Engineering and Computer Science, University of Ljubljana, Slovenia.
  Thesis: Handling noise in inductive logic programming.
• 1995. PhD in Computer Science.
  Faculty of Electrical Engineering and Computer Science, University of Ljubljana, Slovenia.
  Thesis: Numerical constraints and learnability in inductive logic programming.

• Jozef Stefan Institute, Ljubljana, Slovenia.
  1989-1996: Research assistant.
  Since 1996: Research fellow.

• The Turing Institute, Glasgow, Scotland.
  1 November 1991 - 1 July 1992, Supported by a British Council scholarship.
• Institute of Industrial Chemistry, University of Padova, Padova, Italy.
  26 April - 28 May 1993, Supported by TEMPUS JEP 4724.
• Computer Science Department, Katholieke Universiteit Leuven, Leuven, Belgium.
  1 October - 31 December 1993, Supported by a PECO Fellowship from the Commission of the European Communities. Also 26 April -- 26 May 1994.
• Institute for Applied Information Technology, German National Research Center for Computer Science, Sankt Augustin, Germany.
  1 May 1995 - 30 April 1996, Supported by an ERCIM Fellowship.
• Institute of Computer Science, Foundation for Research and Technology-Hellas (FORTH) Science and Technology Park of Crete, Heraklion, Greece.
  1 May 1996 - 31 October 1996, Supported by an ERCIM Fellowship.

• 1985. Third prize at the XXVI International Mathematical Olympiad, Helsinki, Finland.
• 1991. Best paper award at the Third Scandinavian Conference on Artificial Intelligence, Roskilde, Denmark. (With Nada Lavrac, Vladimir Pirnat and Viljem Krizman.)
• 1996. Jozef Stefan Golden Emblem Prize for outstanding doctoral dissertation.

Events

• MLKDD-98. I Brazilian School on Machine Learning and KDD. Rio de Janeiro, September-October 1998. Invited speaker.
• EAIA-98. International Summer School on KDD and Data Mining: Methods and Applications. Caminha, Portugal, September-October 1998. Invited speaker.
• ALT-98. Ninth International Workshop on Algorithmic Learning Theory. Otzenhausen, Germany, October 1998. PC member.
• AIRIES-98. Artificial Intelligence Research In Environmental Science. Victoria, BC, Canada, October 1998. Invited speaker.
• ILP-99. Ninth International Workshop on Inductive Logic Programming. Bled, Slovenia, June 1999. Co-chair with P. Flach.
• ICML-99. Sixteenth International Conference on Machine Learning. Bled, Slovenia, June 1999. Co-chair with I. Bratko.

• IJCAI-95. Fourteenth International Joint Conference on Artificial Intelligence. Montreal, Canada, August 1995. PC member.
• ILP-95. Fifth International Workshop on Inductive Logic Programming. Leuven, Belgium, September 1995. PC member.
• ILP for KDD-96. MLNET Workshop on Data Mining with Inductive Logic Programming. Bari, Italy, July 1996. Member of organizing committee.
• ILP-96. Sixth International Workshop on Inductive Logic Programming. Stockholm, Sweden, August 1996. PC member.
• ECML-97. Ninth European Conference on Machine Learning. Prague, Czech Republic, April 1997. PC member.
• IJCAI Workshop on Frontiers of Inductive Logic Programming. Nagoya, Japan, August 1997. Member of organizing committee.
• ILP&KDD-97. Summer School on Inductive Logic Programming and Knowledge Discovery in Databases. Prague, Czech Republic, September 1997. Program co-chair with N. Lavrac.
• ILP-97. Seventh International Workshop on Inductive Logic Programming. Prague, Czech Republic, September 1997. Program co-chair with N. Lavrac.
• CompulogNet. Area Meeting on Computational Logic and Machine Learning. Manchester, UK, June 1998. Invited speaker.
• ILP-98. Eighth International Conference on Inductive Logic Programming. Madison, Wisconsin, July 1998. PC member.

Machine learning systems

• DINUS
  An ILP system. Successor of LINUS. Learns determinate logic programs.
• LINUS
  With M. Grobelnik and N. Lavrac. An ILP system. Learns constrained logic programs.
• LAGRANGE
  With L. Todorovski. A machine discovery system.
  Learns algebraic, difference, or differential equations linear in the parameters.
• LAGRAMGE
  With L. Todorovski. A machine discovery system.
  Learns nonlinear equations using background knowledge and declarative bias.
• mFOIL
  An ILP systems designed to handle noisy examples.

Memberships

• Journal of Data Mining and Knowledge Discovery. Member of the editorial board.
• ILP Newsletter. Electronic newsletter of the European Scientific Network on Inductive Logic Programming. Co-editor with N. Lavrac.
• Slovenian Artificial Intelligence Society. Board member. Secretary and treasurer (1996-1998).

Projects

• 1989-1992. Information systems and artificial intelligence. National.
• 1989-1992. Expert systems in medicine. National.
• 1989-1992. ESPRIT II BRA Project 3059 ECOLES, Development of representation for machine learning. International.
• 1991-1993. Expert systems -- knowledge-based systems. National.
• 1991-1994. Knowledge engineering. National.
• 1992-1994. Automated knowledge synthesis. National.
• 1992-1995. ESPRIT III BR Project 6020 ILP, Inductive Logic Programming.
  International.
• 1995-1997. Evolutionary Computation in Optimization and System Identification. National.
• 1995-1997. Analysis of clinical databases and synthesis of medical knowledge.. National.
• 1996-1998. ESPRIT IV Project 20237 ILP2, Inductive Logic Programming II.
  International.
• 1996-1999. Inductive Logic Programming and Knowledge Discovery in Ecological Databases. National. Principal investigator.

• 1991-1998. ESPRIT III Network of Excellence in Computational Logic.
  International.
• 1992-1997. ESPRIT III Network of Excellence in Machine Learning. International.
• 1993-1996. PECO Pan-European Scientific Network on
  Inductive Logic Programming. International.
• 1992-1995. TEMPUS JEP 4724 Postgraduate Study of
  Water Resources Management and Sanitary Engineering. International.
• 1996-1998. ESPRIT IV Network of Excellence in Evolutionary Computing.
  International.
• 1998-2001. COPERNICUS Network of Excellence in
  Inductive Logic Programming. International. Scientific coordinator.

Publications (as of 31 December 1996)

• S. Dzeroski. Inductive logic programming and knowledge discovery in databases. In U. Fayyad, G. Piatetsky-Shapiro, P. Smyth, and R. Uthurusamy, editors, Advances in Knowledge Discovery and Data Mining, pages 118--152, MIT Press, Cambridge, MA, 1996.
• S. Dzeroski and I. Bratko. Applications of inductive logic programming. In L. De Raedt, editor, Advances in Inductive Logic Programming, pages 65--81, IOS Press, Amsterdam, 1996.
• S. Dzeroski, J. Grbovic, and W. J. Walley. Machine learning applications in biological classification of river water quality. In R. S. Michalski, I. Bratko, and M. Kubat, editors, Machine Learning, Data Mining and Knowledge Discovery: Methods and Applications. John Wiley and Sons, Chichester, 1997.
• S. Dzeroski and N. Lavrac. Refinement graphs for FOIL and LINUS. In S. H. Muggleton, editor, Inductive Logic Programming, pages 319--333, Academic Press, London, 1992.
• S. Dzeroski, S. Muggleton and S. Russell. PAC-learnability of constrained nonrecursive logic programs. In T. Petsche, S. Judd and S. Hanson, editors, Computational Learning Theory and Natural Learning Systems, Vol. 3, pages 243--255, MIT Press, Cambridge, MA, 1995.
• S. Dzeroski, S. Schulze-Kremer, K. Heidtke, K. Siems, and D. Wettschereck. Diterpene structure elucidation from 13C NMR spectra with machine learning. In N. Lavrac, E. Keravnou and B. Zupan, editors, Intelligent Data Analysis in Medicine and Pharmacology, Kluwer Academic Publishers, Dordrecht, 1997.
• N. Lavrac and S. Dzeroski. Inductive learning of relations from noisy examples. In S. H. Muggleton, editor, Inductive Logic Programming, pages 495--516, Academic Press, London, 1992.
• N. Lavrac, S. Dzeroski, and I. Bratko. Imperfect data handling in inductive logic programming. In L. DeRaedt, editor, Advances in Inductive Logic Programming, pages 48--64, IOS Press, Amsterdam, 1996.
• N. Lavrac, D. Gamberger, and S. Dzeroski. Noise elimination applied in early diagnosis of rheumatic diseases. In N. Lavrac, E. Keravnou and B. Zupan, editors, Intelligent Data Analysis in Medicine and Pharmacology, Kluwer Academic Publishers, Dordrecht, 1997.

• I. Bratko and S. Dzeroski. Engineering applications of inductive logic programming. New Generation Computing, 13: 313--333, (Special issue on Inductive Logic Programming), 1995.
• L. De Raedt and S. Dzeroski. First order jk-clausal theories are PAC-learnable. Artificial Intelligence, 70: 375--392, 1994.
• S. Dzeroski. Learning qualitative models with inductive logic programming. Informatica, 16(4): 30--41, 1992.
• S. Dzeroski, B. Cestnik and I. Petrovski. Using the m-estimate in rule induction. Journal of Computing and Information Technology, 1(1): 37--46, 1993.
• S. Dzeroski, J. Grbovic, W. J. Walley and B. Kompare. Using machine learning techniques in the construction of models. Part II: Rule induction. Ecological Modelling, 1996. In press.
• S. Dzeroski and N. Lavrac. Rule induction and instance-based learning applied in medical diagnosis. Technology and Health Care, 4: 203--221, 1996.
• S. Dzeroski and N. Lavrac. Inductive learning in deductive databases. IEEE Transactions on Knowledge and Data Engineering 5(6): 939--949 (Special issue on Learning and Discovery in Knowledge-Based Databases), 1993.
• S. Dzeroski and L. Todorovski. Discovering dynamics: from inductive logic programming to machine discovery. Journal of Intelligent Information Systems, 4: 89--108 (Special issue on Knowledge Discovery in Databases), 1995.
• J. U. Kietz and S. Dzeroski. Inductive logic programming and learnability. SIGART Bulletin 5(1): 22--32 (Special issue on Inductive Logic Programming), 1994.
• B. Kompare, I. Bratko, F. Steinman and S. Dzeroski. Using machine learning techniques in the construction of models. Part I: Introduction. Ecological Modelling, 75/76: 617--628, 1994.
• V. Krizman, S. Dzeroski, and B. Kompare. Discovering dynamics from measured data. Electrotechnical Review, 62(3-4): 191--198, 1995.
• N. Lavrac and S. Dzeroski. Weakening the language bias in LINUS. Journal on Experimental and Theoretical Artificial Intelligence 6(1): 95--119 (Special issue on Algorithmic Learning Theory), 1994.
• N. Lavrac, S. Dzeroski, V. Pirnat and V. Krizman. The use of background knowledge in learning medical diagnostic rules. Applied Artificial Intelligence, 7(3): 273--293, 1993.
• L. Todorovski and S. Dzeroski. Modeling dynamic systems with machine discovery. Electrotechnical Review, 61(1-2): 55--64, 1994. In Slovenian.
• N. Lavrac, D. Zupanic, I. Weber, D. Kazakov, O. Stepankova, and S. Dzeroski. ILPNET repositories on WWW: Inductive Logic Programming systems, datasets and bibliography. AI Communications, 1996.

• I. Bratko and S. Dzeroski. New AI techniques applied to modelling. In Proc. Second International Conference Design to Manufacture in Modern Industry, pages 359--372, University of Maribor, 1995.
• L. De Raedt, N. Lavrac and S. Dzeroski. Multiple predicate learning. In Proc. Thirteenth International Joint Conference on Artificial Intelligence, pages 1037-1042, Morgan Kaufmann, San Mateo, CA, 1993.
• S. Dzeroski. Handling imperfect data in inductive logic programming. In Proc. Fourth Scandinavian Conference on Artificial Intelligence, pages 111--125, IOS Press, Amsterdam, 1993.
• S. Dzeroski. Learning first-order clausal theories in the presence of noise. In Proc. Fifth Scandinavian Conference on Artificial Intelligence, pages 51--60, IOS Press, Amsterdam, 1995.
• S. Dzeroski, B. Cestnik and I. Petrovski. The use of Bayesian probability estimates in rule induction. In Proc. First Electrotechnical and Computer Science Conference, Volume B, pages 155--158, Slovenia Section IEEE, Ljubljana, Slovenia, 1992.
• S. Dzeroski, L. Dehaspe, B. Ruck and W. Walley. Classification of river water quality data using machine learning. In P. Zannetti, editor, Computer Techniques in Environmental Studies V (Proc. Fifth International Conference on the Development and Application of Computer Techniques to Environmental Studies), Vol. I: Pollution modelling, pages 129--137, Computational Mechanics Publications, Southampton, 1994.
• S. Dzeroski and B. Dolsak. A comparison of relation learning algorithms on the problem of finite element mesh design. In Proc. XXVI Yugoslav Conference of the Society for ETAN, pages 313--320, Ohrid, Macedonia, 1991. In Slovenian.
• S. Dzeroski and J. Grbovic. Knowledge discovery in a water quality database. In Proc. First International Conference on Knowledge Discovery and Data Mining, pages 81--86. AAAI Press, Menlo Park, CA, 1995.
• S. Dzeroski, J. Grbovic and D. Lican-Milosevic. Analysis of water quality data with machine learning. In Proc. Third Electrotechnical and Computer Science Conference, Volume B, pages 175--178, Slovenia Section IEEE, Ljubljana, Slovenia, 1994. In Slovenian.
• S. Dzeroski and N. Lavrac. Learning relations from noisy examples: an empirical comparison of LINUS and FOIL. In Proc. Eighth International Workshop on Machine Learning, pages 399--402, Morgan Kaufmann, San Mateo, CA, 1991.
• S. Dzeroski and D. Lican. Modeling algal growth in the lagoon of Venice with regression trees. In Proc. Second Electrotechnical and Computer Science Conference, Volume B, pages 205--208, Slovenia Section IEEE, Ljubljana, Slovenia, 1993. In Slovenian.
• S. Dzeroski, S. Muggleton and S. Russell. PAC-learnability of determinate logic programs. In Proc. Fifth ACM Workshop on Computational Learning Theory, pages 128--135, ACM Press, New York, 1992.
• S. Dzeroski, S. Muggleton and S. Russell. Learnability of constrained logic programs. In Proc. Sixth European Conference on Machine Learning, pages 342--347, Springer, Berlin, 1993.
• S. Dzeroski and I. Petrovski. Discovering dynamics with genetic programming. In Proc. Seventh European Conference on Machine Learning, pages 347--350, Springer, Berlin, 1994.
• S. Dzeroski, G. Potamias, V. Moustakis, and G. Charissis. Automated revision of expert rules for treating acute abdominal pain in children. In Proc. Sixth European Conference on Artificial Intelligence in Medicine Europe, Springer, Berlin, 1997.
• S. Dzeroski, S. Schulze-Kremer, K. Heidtke, K. Siems, and D. Wettschereck. Applying ILP to diterpene structure elucidation from 13C NMR spectra. In Proc. Sixth International Workshop on Inductive Logic Programming, Springer, Berlin, 1996.
• S. Dzeroski and L. Todorovski. Modeling dynamic systems with machine discovery. In Proc. Second Electrotechnical and Computer Science Conference, Volume B, pages 209--212, Slovenia Section IEEE, Ljubljana, Slovenia, 1993. In Slovenian.
• S. Dzeroski and L. Todorovski. Discovering dynamics. In Proc. Tenth International Conference on Machine Learning, pages 97--103, Morgan Kaufmann, San Mateo, CA, 1993.
• S. Dzeroski and L. Todorovski. Handling real numbers in inductive logic programming. In Proc. Third Electrotechnical and Computer Science Conference, Volume B, pages 143--146, Slovenia Section IEEE, Ljubljana, Slovenia, 1994.
• S. Dzeroski, L. Todorovski and T. Urbancic. Handling real numbers in ILP: a step towards successful behavioral cloning. In Proc. Eighth European Conference on Machine Learning, pages 283--286. Springer, Berlin, 1995.
• B. Kompare and S. Dzeroski. Two artificial intelligence methods for knowledge synthesis from environmental data. In P. Zannetti, editor, Computer Techniques in Environmental Studies V (Proc. Fifth International Conference on the Development and Application of Computer Techniques to Environmental Studies), Vol. II: Environmental Systems, pages 265--272, Computational Mechanics Publications, Southampton, 1994.
• B. Kompare and S. Dzeroski. Getting more out of data: Automated modelling of algal growth with machine learning. In Proc. International Conference on Coastal Ocean Space Utilization, pages 209--220, University of Hawaii, 1995.
• N. Lavrac and S. Dzeroski. Background knowledge and declarative bias in inductive concept learning. In K. Jantke, editor, Proc. Third International Workshop on Analogical and Inductive Inference, pages 51--71, Springer, Berlin, 1992.
• N. Lavrac, S. Dzeroski and M. Grobelnik. Learning nonrecursive definitions of relations with LINUS. In Proc. Fifth European Working Session on Learning, pages 265--281, Springer, Berlin, 1991.
• N. Lavrac, S. Dzeroski, V. Pirnat and V. Krizman. Learning rules for early diagnosis of rheumatic diseases. In Proc. Third Scandinavian Conference on Artificial Intelligence, pages 138--149, IOS Press, Amsterdam, 1991.
• D. Gamberger, N. Lavrac, and S. Dzeroski. Noise elimination in inductive concept learning: A case study in medical diagnosis. In Proc. Seventh International Workshop on Algorithmic Learning Theory, pages 199--212, Springer, Berlin, 1996.
• W. J. Walley and S. Dzeroski. Biological monitoring: A comparison between Bayesian, neural and machine learning methods of water quality classification. In Proc. International Symposium on Environmental Software Systems. Malvern, PA, 1995.
• S. Wrobel and S. Dzeroski. The ILP description learning problem: Towards a general model-level definition of data mining in ILP. In Beitrage zum 7. Fachgruppentreffen Machinelles Lernen der GI-Fachgruppe 1.1.3, pages 33--39, University of Dortmund, 1995.
• B. Zupan and S. Dzeroski. Acquiring and validating background knowledge for machine learning using function decomposition. In Proc. Sixth European Conference on Artificial Intelligence in Medicine Europe, Springer, Berlin, 1997.