Keywords
contingency tables
model selection methods
modelos gráficos
tablas de contingencia
métodos de selección de modelos
How to Cite
Abstract
Loglinear models are frequently utilized in the analysis of association patterns between variables in a multidimensional contingency table. Discrete graphical models are fashionable. They are a subset of the hierarchical loglinear models and its conditional independence structures are represented by a graph. Because of that they can be more easily interpreted. In this work four models selection methods are described and compared: stepwise selection (backward, forward and in two steps) and the EH procedure.A strategy for model selection is proposed.
References
Alfonso, K. (1999) Análisis de Tablas de Contingencia Vía Modelos Gráficos. Tesis de Licenciatura en Matemática, Universidad de La Habana.
Acid, S.; Campos, L.M. (1999) “Fast algorithms for learning simplified graphical models”, in: Proceedings of the Second Symposium on Artificial Intelligence. Editorial Academia, La Habana.
Agresti, A. (1990) Categorical Data Analysis. Wiley, New York.
Asmussen, S.; Edwards, D. (1983) “Collapsability and response variables in contingency tables”, Biometrika 70(3): 566–578.
Darroch, J. N.; Lauritzen, S. L.; Speed, T. P. (1980) “Markov fields and log-linear models for contingency tables”, Annals of Statistics 8: 522–539.
Edwards, D. (1990) “Hierarchical interaction models (with discussion)”, J.Royal Stat. Soc. B 52: 3–20, 51-72.
Edwards, D. (1995a) Introduction to Graphical Modelling. Springer Texts in Statistics, New York.
Edwards, D. (1995b) “Graphical Modelling”, in: J. Krazanowski (Ed.) Recent Advances in Descriptive Multivariate Analysis. Clarendon Press, Oxford.