Showing papers on "Graphical model published in 1987"
Journal Article•
TL;DR: In this article, a strategy for model check- ing based exclusively on exact tests for zero partial association, performed initially in the full table and subsequently in appropriate marginal tables, is presented.
Abstract: The tests for zero partial association in a multidimensional contingency table (suggested by Birch (1965) and later used by Wermuth (1976) for a more elaborate model- searching strategy) have gained new importance with the introduction of the class of graphical models. It is shown how these may be performed as exact conditional tests using as test criteria either the ordinary likelihood ratio, the standard x2 statistic or any other appropriate statistic (e.g. incorporating ordinal properties of some of the categorical variables). A strategy for model check- ing based exclusively on exact tests for zero partial association, performed initially in the full table and subsequently in appropriate marginal tables, is presented. An example of application of the strategy is given, which illustrates that the usual approximate p-values are completely unreliable for large sparse tables.
127 citations
01 Jan 1987
TL;DR: The maximum entropy method is used to reconstruct clutter probability density functions from observed statistical properties; this leads to representations of clutter in terms of Markov random elds (MRF), and it is shown how the set of statistics which is used for each clutter type may be optimised in order to yield a more compact probabilistic model.
Abstract: A thorough understanding of clutter statistics is a prerequisite for the successful analysis of radar images. Usually very simple statistics such as moments and correlation properties are used, perhaps based on an underlying physical model of the scattering and imaging process. In this paper we use the maximum entropy method to reconstruct clutter probability density functions (PDF) from observed statistical properties; this leads to representations of clutter in terms of Markov random elds (MRF). Furthermore we show how the set of statistics which is used for each clutter type may be optimised in order to yield a more compact probabilistic model. The principal advantage of our results is that MRF clutter models may be mapped directly onto parallel image processing hardware, and they provide a rigorous framework for Bayesian decision making concerning the presence of objects embedded in clutter. Image segmentation is another very useful application of these MRF models.
3 citations
2 citations
TL;DR: Two-region saturation at high injection levels and at low frequencies may be conveniently approximated using a graphical model which introduces the concept of quasi-saturation resistance R/SUB QS.
Abstract: Two-region saturation at high injection levels and at low frequencies may be conveniently approximated using a graphical model which introduces the concept of quasi-saturation resistance R/SUB QS/. It is shown how this graphical model may be implemented in the modeling program SPICE with the aid of a distributed equivalent circuit.
1 citations
TL;DR: The research situation (kind of variables and theoretical relations between them) and good presentation and interpretability of graphical models are linked together.
Abstract: Information available should always be considered in any statistical analysis. But it is also true that the good presentation and interpretability of the results should be aimed at. In this paper the research situation (kind of variables and theoretical relations between them) and good presentation and interpretability of graphical models are linked together.