Publications of Dorian
Šuc
Here are some selected publications until 2007. See also my page about
QUIN and Qualitatively Faithful Numerical Learning, and page about
human skill
reconstruction, behavioural cloning and dynamic system
control by ML.
Book
Šuc, D., Vladušič, D., Bratko, I. Qualitatively faithful
quantitative prediction. Artificial
Intelligence, vol 158, no.2, 2004. pp. 189-214, 2004, ISSN 0004-3702. Preprints.
Download at Science
Direct
Bratko, I.,
Šuc, D. Learning qualitative
models. AI Magazine. vol 24, no. 4, pp.
107-119, 2004.
Link to AI Magazine.
Bratko, I.,
Šuc, D. Qualitative data mining and its applications. CIT. J. Comput. Inf. Technol., 2003,
vol. 11, no. 3, pp. 145-150. ISSN 1330-1136.
Link to CIT.
Šuc,
D., Bratko, I. Skill modeling through symbolic
reconstruction of operator's trajectories. IEEE
Transactions on Systems, Man, and
Cybernetics, Part A, Syst. humans, 2000, vol. 30, no. 6, pg.
617-624. ISSN 1083-4427.
Link to IEEE
Transactions at DBLP
Šuc, D., Bratko, I.
Qualitative trees applied to bicycle riding. Electronic Transactions on
Artificial Intelligence, 2000, vol. 4,
Section B, pp. 125-140.
Download at
Linköping
University Electronic Press.
Šuc,
D., Bratko, I. Symbolic and qualitative reconstruction of
control skill.
Electronic Transactions on
Artificial Intelligence, 1999, vol. 3, Section B, pp. 1-22.
Download at
Linköping University
Electronic Press
Book
Chapters, Invited lectures
Bratko, I.,
Šuc, D. Understanding control strategies. In: Della Riccia,
G. (ed.), Dubois, D. (ed.), Kruse, R. (ed.), Lenz, H. (ed.). Planning
based on decision theory, (CISM courses and lectures, no. 472). Wien; New York:
Springer, cop. 2003, pp. 85-98. ISBN 3-211-40756-1.
Bratko, I.,
Šuc, D. Qualitative data mining and its applications. In: ITI 2003 :
proceedings of the 25th International Conference on Information Technology
Interfaces, June 16-19, 2003,
Cavtat,
Croatia. Zagreb: University
of Zagreb, SRCE University
Computing Centre, 2003, pp. 3-8.
Bratko, I.,
Šuc, D. Qualitative explanation of controllers. In:
Sixteenth
International Workshop on Qualitative Reasoning QR 2002, June 10-12, 2002, Sitges
- Barcelona - Catalonia
- Spain.
Sevilla: Edición Digital, 2002, pp. 1-2.
Bratko, I., Šuc, D. Qualitative data mining. In:
The new trends in knowledge processing data mining, semantic web and
computational science : the sixth SANKEN international symposium, 2003, pp.
10-12.
Selected Published Scientific
Conference Contributions
Mele. K., Maver, J., Šuc, D. Image Categorization Using Local Probabilistic
Descriptors, Proceedings of the 18th International Conf. on Pattern
Recognition, ICPR 2006, (Extended
version of the published paper).
Žabkar, J., Vladušič, D., Žabkar, R., Čemas, D., Šuc,
D., Bratko, I. Using Qualitative Constraints In Ozone Prediction, Proceedings of the 19th International Workshop on Qualitative
Reasoning, 2005.
Šuc, D., Bratko, I. Combining Learning
Constraints and Numerical Regression. Proceedings
of 19th International Joint Conference on Artificial Intelligence
IJCAI-05, 2005.
Šuc, D., Bratko, I., Sammut,
Claude. Learning to Fly Simple and Robust. Machine learning : ECML 2004,
Proceedings of the 15th European Conf. on
Machine Learning, pp. 407-418.
Link to electronic edition.
Šuc, D., Vladušič,
D., Bratko, I. Qualitatively
faithful quantitative prediction. Proceedings of the eighteenth International
Joint Conference on Artificial Intelligence, pp. 1052-1057, San Francisco: Morgan
Kaufmann Publishers, 2003. Acapulco,
August, 2003.
Šuc, D., Bratko, I. Improving
numerical prediction with qualitative constraints. Machine learning :
ECML 2003, Proceedings of th 14th European
Conf. on Machine Learning, (Lecture notes in computer science, Lecture
notes in artificial intelligence, vol. 2837), , pp. 385-396. Berlin;
Heidelberg; New York: Springer, cop. 2003.
Bratko, I., Šuc, D. Using
machine learning to understand operator's skill. Developments in
applied artificial intelligence : proceedings, (Lecture notes in computer
science, Lecture notes in artificial intelligence, 2358). pp. 812-823. Berlin [etc.]: Springer,
cop. 2002.
Šuc, D., Bratko, I. Qualitative reverse engineering. In: Machine learning : proceedings of the
19th International Conference (ICML 2002), University
of New South Wales, Sydney, Australia,
July 8-12, 2002. San Francisco:
Morgan Kaufmann, 2002, pp. 610-617. Slides in PowerPoint.
Bandelj, A., Bratko, I.,
Šuc, D. Qualitative
simulation with CLP. In Proceedings of the Sixteenth International
Workshop on Qualitative Reasoning, (QR2002), Sitges, Spain, 2002, pp. 5-9.
Šuc, D., Bratko, I.
Qualitative induction.
In: Fifteenth International Workshop on Qualitative Reasoning, May 17-18,
2001, St. Mary's University San Antonio,
Texas. Stoughton: The Printing House, 2001, pg.
13-20. Slides in PowerPoint.
Šuc, D. Learning qualitative strategies.
In: IC-AI'2001 : proceedings of the International Conference on Artificial
Intelligence, Las Vegas, Nevada, USA
June 25-28, 2001. [s.l.]: CSREA Press, 2001, vol.
2, pg. 1016-1022.
Šuc, D., Bratko, I. Induction
of qualitative trees. In: 12th European
Conference on Machine Learning, Freiburg, Germany,
2001. Machine learning : ECML 2001 : proceedings, (Lecture notes in
computer science, Lecture notes in artificial intelligence, 2167). Berlin [etc.]: Springer,
2001, pg. 442-453. Slides in PowerPoint.
Šuc, D., Bratko, I.
Qualitative induction
for behavioural cloning. Proceedings
of Electrotechnical and Computer Science Conference ERK 2001,
Slovenia,
Slovene section IEEE, 2001, vol. B, pp. 153-156. Slides in PowerPoint (partially in Slovene). The paper received
award for the best paper based on Ph.D. thesis.
Šuc, D., Bratko, I. Qualitative trees applied to bicycle riding. In:
Furukawa, K. (ed.). Machine intelligence 17 : special focus on life long
learning and discovery in procedural and declarative knowledge. 2000, pp.
81-93.
Šuc, Dorian, Bratko, I.
Problem decomposition
for behavioural cloning. In: ECML 2000: 11th European Conference on Machine
Learning, Machine
learning : ECML 2000 : proceedings, (Lecture notes in computer science,
Lecture notes in artificial intelligence, 1810). Berlin [etc.]: Springer, 2000, pg. 382-391.
Šuc, D., Bratko, I. Modelling of control skill by qualitative constraints.
In: Thirteenth International Workshop on Qualitative
Reasoning, Loch Awe, Scotland, 7-9 June 1999. [Aberystwyth: University
of Aberystwyth,
1999], pg. 212-220.
Šuc, D.,
Bratko, I. Skill modeling through
symbolic reconstruction of operator's trajectories. In: Automated
systems based on human skill : joint design of technology and organisation : preprints of the 6th IFAC Symposium, 1997. Aachen:
University of Technology,
Department of Informatics in Mechanical Engineering; Ljubljana: J. Stefan Institute, 1997, pp.
35-38. See also Skill
moddeling through symbolic reconstruction of
operator's trajectorijes, full version, Techical Report
Šuc, D., Bratko, I.
Skill reconstruction as
induction of LQ controllers with subgoals.
In: IJCAI-97 : proceedings of the fiftheenth
International joint conference on artificial intelligence, Nagiya, Japan August 23-29, 1997. Volume 2.
[S.l.]: International joint conference on artificial
intelligence, 1997, pp. 914-920.
Some Papers in
Slovene
Šuc, D.
Modeliranje
veščine vodenja s simboličnim posploševanjem trajektorij vodenja, (zipped file),
Master Th., Faculty of Computer and Information Sciences, University of Ljubljana, Slovenia,
1998
Šuc, D.
Modeliranje veščine vodenja s simboličnim posploševanjem trajektorij vodenja. In: ZAJC, Baldomir (ed.). Zbornik sedme Elektrotehniške in računalniške
konference ERK '98, 24. - 26. september
1998, Portorož, Slovenija.
Ljubljana: IEEE
Region 8, Slovenska sekcija
IEEE, 1998, vol. A, pg. 413-416.
Šuc, D.
Kvalitativno modeliranje veščine vodenja. (In Slovene)
V: ZAJC, Baldomir (ed.). Zbornik
osme Elektrotehniške in računalniške konference ERK
'99, 23. - 25. september 1999, Portorož,
Slovenija. Ljubljana:
IEEE Region 8, Slovenska sekcija
IEEE, 1999, vol. A, pg. 325-328.
Šuc, D.
Strojno učenje kvalitativnih strategij vodenja. (In Slovene) In: BAVEC, Cene (ed.),
GAMS, Matjaž (ed.). Mednarodna
multi-konferenca Informacijska
družba, Ljubljana,
1999. Informacijska družba
IS'99 : zbornik mednarodne
multi-konference, 12. do 14. oktobra
1999, Ljubljana, Slovenija : proceedings of the
international multi-conference, 12 - 14 October 1999. Ljubljana: Institut
Jožef Štefan, 1999, pg.
68-73.
Thesis (Machine
Reconstruction of Human Control Strategies)
Šuc, D. Machine reconstruction of human control strategies,
Doctoral dissertation, Ljubljana,
2001. VIII, 175 pages, Ph.D.Th.
Slides in PowerPoint. The
dissertation is to be published with IOS Press (in October 2003): Dorian Šuc,
Machine reconstruction of human control strategies, IOS Press, series
Frontiers in Artificial Intelligence and Applications.
Abstract: Complex dynamic systems are usually controlled by
operators who acquired their skill through years of experience. Typically, such
a control skill is sub-cognitive and hard to reconstruct through introspection.
The operators cannot completely describe their skill, but can demonstrate it.
Therefore an attractive approach to the reconstruction of human control skill
involves machine learning from operator's execution traces. The goal is to
induce a model of the operator's skill, a control strategy
that helps to understand the skill and can be used to control the system.
Behavioural cloning is an approach to such skill
reconstruction. In the "original'' approach to behavioural
cloning a strategy is induced as a direct mapping from system's states to
actions in the form of a decision or regression tree. This thesis develops new
ideas to tackle problems that were generally observed with this approach to
human skill reconstruction.
One idea is to decompose the learning problem and induce goal-directed
strategies that consider learned models of system's dynamics. We introduce a
generalized operator's trajectory that can be seen as a continuously changing subgoal. This improves the robustness of the resulting
controllers.
Another idea, that is also relevant to the comprehensibility, is to induce qualitative
models of human control skill. We show that such qualitative strategies provide
an insight into the operator's control skill. On the basis on our experiments,
we believe that qualitative strategies can capture important and non-trivial
aspects of human control skill. Qualitative strategies open also other new
perspectives to the reconstruction of human control skill, such as
reconstruction of individual differences in operators's
control styles.
These ideas were implemented and evaluated in dynamic domains including
container crane, a double pendulum referred to as the acrobot,
and bicycle ridding. To induce qualitative control strategies we developed
program QUIN for learning qualitative constraint trees from numerical examples.
Keywords: artificial intelligence, machine learning, qualitative modelling, behavioural cloning,
skill reconstruction, human control strategies, dynamic systems, system
control.