MILAN STUDENY - HOME PAGE

Affiliation

Milan Studený

Institute of Information Theory and Automation

Academy of Sciences of the Czech Republic

Pod vodarenskou vezi 4

18208 Prague 8 Liben

Czech Republic

Home mail address

Bryksova 951

19800 Prague 14 Cerny Most II

Czech Republic

Other connection

Phone (job): (420-) 2 6605 2304

Fax: (420-) 2 688 4903 (institute)

Fax: (420-) 2 6897 008 (department)

E-mail: studeny@utia.cas.cz

Home phone: (420-) 2 81911 553

Research interests

formal properties of conditional independence (also in different calculi of AI)
using graphs for modelling conditional independence structure (Bayesian networks, chain graphs)
probabilistic knowledge based systems
general questions of probability theory
marginal problem
learning Bayesian networks

Some web references

the third European Workshop Probabilistic Graphical Models (PGM06), to be held in September 12-15, Prague, Czech Republic.
HSSS research kitchen Learning Conditional Independence Models October 16-20, 2000, Trest, Czech Republic.

Selected publications (updated on June 12, 2009)

Axiomatic properties of conditional independence

M. Studeny: Multiinformation and the problem of characterization of conditional independence relations. Problems of Control and Information Theory 18 (1989), n. 1, pp. 3-16.
For an abstract click here. No postscript version is available. To get a hard copy of the paper please contact the author at studeny@utia.cas.cz.
M. Studeny: Attempts at axiomatic description of conditional independence. Kybernetika 25 (1989), supplement to n. 3, pp. 72-79.
For an abstract click here. No postscript version is available. Only a few hard copies left. To get a copy of the paper please contact the author at studeny@utia.cas.cz .
M. Studeny: Conditional independence relations have no finite complete characterization. In Information Theory, Statistical Decision Functions and Random Processes. Transactions of the 11th Prague Conference vol. B (S. Kubik, J.A. Visek eds.) Kluwer, Dordrecht - Boston - London (also Academia, Prague) 1992, pp. 377-396.
For an abstract click here. A postscript version (349kB) is available.
M. Studeny: Semigraphoids are two-antecedental approximations of stochastic conditional independence models. In Uncertainty in Artificial Intelligence. Proceedings of the Tenth Conference (R.L. de Mantaras, D. Poole eds.) Morgan Kaufmann, San Francisco CA 1994, pp. 546-552.
For an abstract click here. A postscript version (183kB) is available.
M. Studeny: Semigraphoids and structures of probabilistic conditional independence. Annals of Mathematics and Artificial Intelligence 21 (1997), n. 1, pp. 71-98. For an abstract click here. A postscript version (315 kB) is available.

Description of conditional independence structures

M. Studeny: Structural semigraphoids. International Journal of General Systems 22 (1994), n. 2, pp. 207-217. For an abstract click here. A postscript version (223kB) is available.
M. Studeny: Description of structures of stochastic conditional independence by means of faces and imsets. 1st part: introduction and basic concepts. International Journal of General Systems 23 (1994), n. 2, pp. 123-137. For an abstract click here. No postsript version is available. To get a hard copy of the paper please contact the author at studeny@utia.cas.cz . On the other hand, the content of the paper is covered by later DrSc thesis mentioned below.
M. Studeny: Description of structures of stochastic conditional independence by means of faces and imsets. 2nd part: basic theory. International Journal of General Systems 23 (1995), n. 3, pp. 201-219. For an abstract click here. No postsript version is available. To get a hard copy of the paper please contact the author at studeny@utia.cas.cz . On the other hand, the content of the paper is covered by later DrSc thesis mentioned below.
M. Studeny: Description of structures of stochastic conditional independence by means of faces and imsets. 3rd part: examples of use and appendices. International Journal of General Systems 23 (1995), n. 4, pp. 323-341. For an abstract click here. No postsript version is available. To get a hard copy of the paper please contact the author at studeny@utia.cas.cz . On the other hand, the content of the paper is covered by later DrSc thesis mentioned below.
M. Studeny: Description of conditional independence structures by means of imsets: a connection with product formula validity. In Uncertainty in Intelligent Systems (B. Bouchon-Meunier, L. Valverde, R.R. Yager eds.) North-Holland, Amsterdam - London - New York - Tokyo 1993, pp. 179-194.
For an abstract click here. A postscript version (277kB) is available.
M. Studeny: Complexity of structural models. In Prague Stochastics'98. Proceedings of the Joint Session of the 6th Prague Conference on Asymptotic Statistics and the 13th Prague Conference on Information Theory, Statistical Decision Functions and Random Processes vol. II, August 23-28, 1998, Prague, Czech Republic, pp. 521-528.
For an abstract click here. A postscript version (170kB) is available.
M. Studeny: On non-graphical description of models of conditional independence structure. Paper presented at HSSS Workshop on Stochastic Systems for Individual Behaviours, January 22-23, 2001, Louvain la Neuve, Belgium (15 pages).
For an abstract click here. A postscript version (493kB) is available.
M. Studeny: On mathematical description of probabilistic conditional independence structures, thesis for DrSc degree, Institute of Information Theory and Automation, Academy of Sciences of the Czech Republic, Prague, May 2001, 192 pages.
For an abstract click here. A postscript version (3284kB) is available. WARNING: the file is very long, it has 192 pages and contains some colorful pictures.
Note it the thesis was the basis for later book M. Studeny: Probabilistic Conditional Independence Structures, Springer London 2005 (series Information Science and Statistics).
Contact Springer-Verlag if you wish to buy the book.
M. Studeny: On stochastic conditional independence: the problems of characterization and description. Annals of Mathematics and Artificial Intelligence 35 (2002), pp. 323-241.
For an abstract click here. A postscript version of a preprint (489 kB) is available.
P. Simecek, M. Studeny: Vyuziti pojmu Hilbertovy baze pro overovani hypotezy o shodnosti strukturalnich a kombinatorickych imsetu (in Czech, translation: The use of the concept of a Hilbert basis to verify a hypothesis about the coincidence of structural and combinatorial imsets), in ROBUST 2004 Sbornik praci trinacte letni skoly JCMF (J. Antoch, G. Dohnal eds.), Trest, Czech Republic, JCMF 2004, pp. 395-401. For an abstract click here. A pdf version (178kB) is available.

Conditional independence inference

M. Studeny, R.R. Bouckaert, T. Kocka: Extreme supermodular set functions over five variables. Research report n. 1977, Institute of Information Theory and Automation, Prague, January 2000 (32 pages). For an abstract click here. A postscript version (434kB) is available.
Moreover, the result of the work is also an electronic catalogue of representatives extreme supermodular set functions over five variables.
M. Studeny: Structural imsets: an algebraic method for describing conditional independence structures. In Proceedings of the 10th International Conference IPMU 2004 (B. Bouchon-Meunier, G. Colleti, R. R. Yager eds.), Editrice Universita La Sapienza, Perugia, Italy, 2004, vol. 2, pp. 1323-1330.
For an abstract click here. A postscript version (307kB) is available.
R. R. Bouckaert, M. Studeny: Racing algorithms for conditional independence inference. International Journal of Approximate Reasoning 45 (2007), pp. 386-401. For an abstract click here. A pdf version of a preprint (347kB) is available.

Conditional independence structures over four variables

M. Studeny and P. Bocek: CI-models arising among 4 random variables. In Proceedings of the 3rd workshop WUPES, September 11-15, 1994, Trest, Czech Republic, pp. 268-282.
For an abstract click here. A postscript version (192kB) is available.
F. Matus and M. Studeny: Conditional independences among four random variables I. Combinatorics, Probability and Computing 4 (1995), n. 3, pp. 267-278.
For an abstract click here. This is the first paper of a series of papers written by Frantisek Matus. You may either download ps versions of these papers from here below or contact Frantisek Matus at matus@utia.cas.cz (for Matus's home page of click here):

F. Matus and M. Studeny (1995) Conditional independences among four random variables I. Combinatorics, Probability & Computing 4 269-278.
F. Matus (1995) Conditional independences among four random variables II. Combinatorics, Probability & Computing 4 407-417.
F. Matus (1999) Conditional independences among four random variables III: final conclusion. Combinatorics, Probability & Computing 8 269-276.

Measures of dependence

M. Studeny: Pojem multiinformace v pravdepodobnostnim rozhodovani. (in Czech, translation: The notion of multiinformation in probabilistic decision making), doctoral thesis, Czechoslovak Academy of Sciences, Institute of Information Theory and Automation, Prague 1987, Czechoslovakia, 121 pages.
A postscript abstract (69kB) is available.
M. Studeny: Asymptotic behaviour of empirical multiinformation. Kybernetika 23 (1987), n. 2, pp. 124-135.
For an abstract click here. No postscript version is available. Only a few hard copies left . To get a copy of the paper please contact the author at studeny@utia.cas.cz .
J. Zvarova and M. Studeny: Information theoretical approach to constitution and reduction of medical data. International Journal of Medical Informatics 45 (1997), n. 1-2, pp. 65-74.
For an abstract click here. No postscript version is available. To get a hard copy of the paper please contact Jana Zvarova at zvarova@uivt.cas.cz .
M. Studeny and J. Vejnarova: The multiinformation function as a tool for measuring stochastic dependence. In Learning in Graphical Models (M. I. Jordan ed.) Kluwer, Dordrecht 1998, pp. 261-298. For an abstract click here. A postscript version (379kB) is available.

Conditional independence in non-probabilistic calculi

M. Studeny: Formal properties of conditional independence in different calculi of AI. In Symbolic and Quantitative Approaches to Reasoning and Uncertainty (M. Clarke, R. Kruse, S. Moral eds.) Springer-Verlag, Berlin -Heidelberg 1993, pp. 341-348.
For an abstract click here. A postscript version (210kB) is available.
M. Studeny, F. Matus and J. Vejnarova: Decomposition of large systems and independence structures. In Proceedings of the 2nd European Congress on System Science, October 5-8, 1993, Prague, Czech Republic, pp. 891-898. For an abstract click here. No postscript version is available.
M. Studeny: Conditional independence and natural conditional functions. International Journal of Approximate Reasoning 12 (1995), n. 1, pp. 43-68.
For an abstract click here. A postscript version (361kB) is available.

Chain graphs

R. R. Bouckaert and M. Studeny: Chain graphs: semantics and expressiveness. In Symbolic and Quantitative Approaches to Reasoning and Uncertainty (Ch. Froidevaux, J. Kohlas eds.) Lecture Notes in Artificial Intelligence 946 (subseries of Lecture Notes in Computer Science), Springer-Verlag, Berlin - Heidelberg 1995, pp. 69-76.
For an abstract click here. A postscript version (146kB) is available.
M. Studeny: On separation criterion and recovery algorithm for chain graphs. In Uncertainty in Artificial Intelligence. Proceedings of the 12th Conference (E. Horvitz, F. Jensen eds.) Morgan Kaufmann, San Francisco 1996, pp. 509-516.
For an abstract click here. A postscript version (233kB) is available.
M. Studeny: A recovery algorithm for chain graphs. International Journal of Approximate Reasoning 17 (1997), n. 2-3, pp. 265-293.
For an abstract click here. A postscript version (312kB) is available.
M. Studeny and R. R. Bouckaert: On chain graph models for description of conditional independence structures. The Annals of Statistics 26 (1998), n. 4, pp. 1434-1495.
For an abstract click here. A postscript version (566kB) is available.
M. Studeny: Bayesian networks from the point of view of chain graphs. In Uncertainty in Artificial Intelligence. Proceedings of the 14th Conference (G. F. Cooper, S. Moral eds.) Morgan Kaufmann, San Francisco 1998, pp. 496-503.
For an abstract click here. A postscript version (206kB) is available.
M. Volf and M. Studeny: A graphical characterization of the largest chain graphs. International Journal of Approximate Reasoning 20 (1999), pp. 209-236.
For an abstract click here. A postscript version (1319kB) is available.
M. Studeny: O pouziti retezcovych grafu pro popis struktur podminene nezavislosti (in Czech, translation: On the use of chain graphs for the description of conditional independence structures) In ROBUST'2002. Sbornik praci dvacate zimni skoly JCMF, Hejnice, Ceska republika, pp. 292-314.
This is an overview paper in Czech written for students of statistics. The topic of the paper are basic graphical tools for the description of conditional independence structures. For an abstract (in Czech) click here. A postscript version (570kB) is available.
M. Studeny: Characterization of essential graphs by means of the operation of legal merging of components. International Journal of Uncertainty, Fuzziness, and Knowledge-Based Systems 12 (2004), pp. 43-62. For an abstract click here. A postscript version of a preprint (590kB) is available.
A. Roverato, M. Studeny: A graphical representation of equivalence classes of AMP chain graphs. Journal of Machine Learning Research 7 (2006), pp. 1045-1078.
For an abstract click here. A pdf version of a preprint (373kB) is available.
M. Studeny, A. Roverato, S. Stepanova: Two operations of merging and splitting components in a chain graph. Kybernetika 45 (2009), n. 2, pp. 208-248.
For an abstract click here. A pdf version of the published paper (451kB) is available.

Learning Bayesian networks

T. Kocka, R. R. Bouckaert, M. Studeny: On characterizing inclusion of Bayesian networks. In Uncertainty in Artificial Intelligence. Proceedings of the 17th Conference (J. Breese, D. Koller eds.) Morgan Kaufmann, San Francisco 2001, pp. 261-268.
For an abstract click here. A postscript version (425kB) is available.
T. Kocka, R. R. Bouckaert, M. Studeny: On the inclusion problem. Research report n. 2010, Institute of Information Theory and Automation, Prague, February 2001.
For an abstract click here. A postscript version (841kB) is available.
M. Studeny: An algebraic approach to learning Bayesian networks. In the book of abstracts of the 7th International Valencia Meeting on Bayesian Statistics, Playa de las Americas, Tenerife, Spain, June 1-6, 2002, p.179.
For an abstract click here. A postscript version of posters (1165kB) is available.
M. Studeny: Characterization of inclusion neighbourhood in terms of the essential graphs. International Journal of Approximate Reasoning 38 (2005), n. 3, pp. 283-309.
For an abstract click here. A poscript version of a preprint (529kB) is available.
M. Studeny: An algebraic approach to structural learning Bayesian networks. In Proceedings of the 11th International Conference IPMU 2006 (B. Bouchon-Meunier, R. R. Yager eds.), Editions EDK, Paris, France 2006, pp. 2284-2291.
For an abstract click here. A pdf version (176kB) is available.
J. Vomlel, M. Studeny: Graphical and algebraic representatives of conditional independence models. Submitted to Advances in Bayesian Networks (P. Lucas, J. A. Gamez, A. Salmeron eds.), Springer-Verlag 2006. For an abstract click here. A pdf version of a preprint (268kB) is available.
M. Studeny, J. Vomlel: A reconstruction algorithm for the essential graph. International Journal of Approximate Reasoning 50 (2009), pp. 385-413.
For an abstract click here. A pdf version of the preprint (464kB) is available.
M. Studeny: Mathematical aspects of learning Bayesian networks: Bayesian quality criteria. Research report n. 2234, Institute of Information Theory and Automation, Prague, December 2008. For an abstract click here. A pdf version (484kB) is available.

Other graphical approaches

M. Studeny: Comparison of graphical approaches to description of conditional independence structures. In Proceedings of WUPES`97, January 22-25, 1997, Prague, Czech Republic, pp. 156-172. For an abstract click here. A postscript version (208kB) is available.
A. Paz, R. Y. Geva and M. Studeny: Representation of irrelevance relations by annotated graphs. Fundamenta Informaticae 42 (2000), pp. 149-199.
For an abstract click here. A postscript version (498kB) is available.

Marginal problem

M. Studeny: Marginal problem in different calculi of AI. In Advances in Intelligent Computing - IPMU'94 (B. Bouchon-Meunier, R. R. Yager, L. A. Zadeh eds.), Lecture Notes in Computer Science 945, Springer-Verlag, Berlin - Heidelberg 1995, pp. 348-359.
For an abstract click here. A postscript version (192kB) is available.
M. Studeny: On marginalization, collapsibility and precollapsibility. In Distributions with Given Marginals and Moment Problems (V. Benes, J. Stepan eds.) Kluwer, Dordrecht 1997, pp. 191-198. For an abstract click here. A postscript version (148kB) is available.
A. Perez, M. Studeny: Comparison of two methods for approximation of probability distributions with prescribed marginals. Kybernetika 43 (2007), n. 5, pp. 591-618. For an abstract click here. A pdf version of the published paper (462kB) is available.

Differentiation of measures

M. Studeny: Derivovani mer v metrickych prostorech. (in Czech, translation: differentiation of measures in metric spaces), graduate diploma thesis, Faculty of Mathematics and Physics, Charles University, Prague 1981, Czechoslovakia, 115 pages.
A postscript abstract (58kB) is available. No postsript version is available.
M. Studeny: On the differentiation theorem in metric groups. Commentationes Mathematicae Universitatis Carolinae 24 (1983), n. 2, pp. 223-232. For an abstract click here. A scanned pdf copy (826kB) is available.

Other topics

M. Studeny: Convex cones in finite-dimensional real vector spaces. Kybernetika 29 (1993), n. 2, pp. 180-200.
For an abstract click here. A postscript version (294kB) is available.
P. Lachout, M. Studeny and J. Sindelar: On set-valued measures. Informatica 4 (1993), n. 1-2, pp. 21-44.
For an abstract click here. No postsript version available. To get a hard copy of the paper please contact either Petr Lachout at lachout@karlin.mff.cuni.cz or Jan Sindelar at sindelar@utia.cas.cz .