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.

Complexity of a model of conditional independence structure is introduced as the least cardinality of a generating set. Four basic types of complexity are distinguished which depend on the type of generating. A method of calculation of complexity of a given conditional independence model is proposed. The method is based on a more effective way of representation of the model by means of a list of dominant conditional independence statements. Prospects of the proposed approach are discussed in the end.

AMS classification 68T30, 62H05

probabilistic model
dominant triplet

A pdf copy (converted postscript version) (185kB) is available.

The paper partially builds on the work: