Geodesic
A comparison of evidential networks and compositional models
Kybernetika, Tome 50 (2014) no. 2, pp. 246-267
Cet article a éte moissonné depuis la source Czech Digital Mathematics Library

Voir la notice de l'article

Several counterparts of Bayesian networks based on different paradigms have been proposed in evidence theory. Nevertheless, none of them is completely satisfactory. In this paper we will present a new one, based on a recently introduced concept of conditional independence. We define a conditioning rule for variables, and the relationship between conditional independence and irrelevance is studied with the aim of constructing a Bayesian-network-like model. Then, through a simple example, we will show a problem appearing in this model caused by the use of a conditioning rule. We will also show that this problem can be avoided if undirected or compositional models are used instead.
Several counterparts of Bayesian networks based on different paradigms have been proposed in evidence theory. Nevertheless, none of them is completely satisfactory. In this paper we will present a new one, based on a recently introduced concept of conditional independence. We define a conditioning rule for variables, and the relationship between conditional independence and irrelevance is studied with the aim of constructing a Bayesian-network-like model. Then, through a simple example, we will show a problem appearing in this model caused by the use of a conditioning rule. We will also show that this problem can be avoided if undirected or compositional models are used instead.
DOI : 10.14736/kyb-2014-2-0246
Classification : 62H17, 62H99, 68T37
Keywords: evidence theory; conditioning; independence; directed graphs 
@article{10_14736_kyb_2014_2_0246,
     author = {Vejnarov\'a, Ji\v{r}ina},
     title = {A comparison of evidential networks and compositional models},
     journal = {Kybernetika},
     pages = {246--267},
     year = {2014},
     volume = {50},
     number = {2},
     doi = {10.14736/kyb-2014-2-0246},
     mrnumber = {3216993},
     zbl = {06325223},
     language = {en},
     url = {http://geodesic.mathdoc.fr/articles/10.14736/kyb-2014-2-0246/}
}
TY  - JOUR
AU  - Vejnarová, Jiřina
TI  - A comparison of evidential networks and compositional models
JO  - Kybernetika
PY  - 2014
SP  - 246
EP  - 267
VL  - 50
IS  - 2
UR  - http://geodesic.mathdoc.fr/articles/10.14736/kyb-2014-2-0246/
DO  - 10.14736/kyb-2014-2-0246
LA  - en
ID  - 10_14736_kyb_2014_2_0246
ER  - 
%0 Journal Article
%A Vejnarová, Jiřina
%T A comparison of evidential networks and compositional models
%J Kybernetika
%D 2014
%P 246-267
%V 50
%N 2
%U http://geodesic.mathdoc.fr/articles/10.14736/kyb-2014-2-0246/
%R 10.14736/kyb-2014-2-0246
%G en
%F 10_14736_kyb_2014_2_0246
Vejnarová, Jiřina. A comparison of evidential networks and compositional models. Kybernetika, Tome 50 (2014) no. 2, pp. 246-267. doi: 10.14736/kyb-2014-2-0246

[1] Beeri, C., Fagin, R., Maier, D., Yannakakis, M.: On the desirability of acyclic database schemes. J. Association for Computing Machinery 30 (1983), 479-513. | DOI | MR | Zbl

[2] Yaghlane, B. Ben, Smets, Ph., Mellouli, K.: Belief functions independence: I. The marginal case. Internat. J. Approx. Reasoning 29 (2002), 47-70. | DOI | MR

[3] Yaghlane, B. Ben, Smets, Ph., Mellouli, K.: Belief functions independence: II. The conditional case. Internat. J. Approx. Reasoning 31 (2002), 31-75. | DOI | MR

[4] Yaghlane, B. Ben, Smets, Ph., Mellouli, K.: Directed evidential networks with conditional belief functions. In: Proc. ECSQARU 2003 (T. D. Nielsen and N. L. Zhang, eds.), pp. 291-305. | MR

[5] Benferhat, S., Dubois, D., Gracia, L., Prade, H.: Directed possibilistic graphs and possibilistic logic. In: Proc. IPMU'98 (B. Bouchon-Meunier and R. R. Yager eds.), Editions E.D.K. Paris, pp. 1470-1477.

[6] Couso, I., Moral, S., Walley, P.: Examples of independence for imprecise probabilities. In: Proc. ISIPTA'99 (G. de Cooman, F. G. Cozman, S. Moral, and P. Walley, eds.), pp. 121-130.

[7] Cozman, F. G.: Credal networks. Artificial Intelligence J. 120 (2000), 199-233. | DOI | MR | Zbl

[8] Daniel, M.: Belief conditioning rules for classic belief functions. In: Proc. WUPES'09 (T. Kroupa and J. Vejnarová, eds.), pp. 46-56.

[9] Cooman, G. De: Possibility theory I - III. Internat. J. General Systems 25 (1997), 291-371. | DOI

[10] Guan, J. W., Bell, D. A.: Evidence Theory and its Applications. Vol. 1. North-Holland, 1991. | MR

[11] Jiroušek, R., Vejnarová, J.: Compositional models and conditional independence in evidence theory. Internat. J. Approx. Reasoning 52 (2011), 316-334. | DOI | MR | Zbl

[12] Jiroušek, R., Vejnarová, J., Daniel, M.: Compositional models for belief functions. In: Proc. ISIPTA'07 (G. De Cooman, J. Vejnarová, and M. Zaffalon, eds.) Praha, pp. 243-252.

[13] Kong, A.: Multivariate Belief Functions and Graphical Models. Doctoral disertation, Department of Statistics, Harvard University, 1986.

[14] Shafer, G.: A Mathematical Theory of Evidence. Princeton University Press, Princeton, New Jersey 1976. | MR | Zbl

[15] Shenoy, P. P.: Conditional independence in valuation-based systems. Internat. J. Approx. Reasoning 10 (1994), 203-234. | DOI | MR | Zbl

[16] Studený, M.: Formal properties of conditional independence in different calculi of artificial intelligence. In: Proc. ECSQARU'93 (K. Clarke, R. Kruse, and S. Moral, eds.), Springer-Verlag, 1993, pp. 341-348.

[17] Vejnarová, J.: Conditional independence relations in possibility theory. Internat. J. Uncertainty, Fuzziness and Knowledge-Based Systems 8 (2000), 253-269. | DOI | MR | Zbl

[18] Vejnarová, J.: On conditional independence in evidence theory. In: Proc. ISIPTA'09 (T. Augustin, F. P A. Coolen, S. Moral, M. C. M. Troffaes, eds.), Durham 2009, pp. 431-440.

[19] Vejnarová, J.: An alternative approach to evidential network construction. In: Combining Soft Computing and Statistical Methods in Data Analysis (Ch. Borgelt, G. Gonzales-Rodriguez, W. Trutschnig, M. A. Lubiano, M. A. Gil, P. Grzegorzewski, and O. Hryniewicz, eds.), Oviedo 2010, pp. 619-626.

[20] Vejnarová, J.: Conditioning, conditional independence and irrelevance in evidence theory. In: Proc. ISIPTA'11 (F. Coolen, G. de Cooman, T. Fetz, and M. Oberguggenberger, eds.), Innsbruck 2011, pp. 381-390.

[21] Vejnarová, J.: Conditioning in evidence theory from the perspective of multidimensional models. In: Proc. IPMU'12 (S. Greco et al., eds.), Part III, CCIS 299, 2012, pp. 450-459. | Zbl

[22] Vejnarová, J.: Evidential networks from a different perspective. In: Synergies of Soft Computing and Statistics for Intelligent Data Analysis, Soft Methods In Probability and Statistics, Heidelberg 2012, pp. 429-436.

[23] Vejnarová, J.: On weakness of evidential networks. In: Proc. 9th Workshop on Uncertainty Processing, pp. 190-203.

[24] Xu, H., Smets, Ph.: Evidential reasoning with conditional belief functions. In: Proc. Tenth Conference on Uncertainty in Artificial Intelligence (UAI'94), pp. 598-605.

Cité par Sources :