In this light, I’m not sure that your challenge makes sense. Bayes nets are a way of expressing a properties of probabilistic models, rather than a model unto themselves.
Valid point, but I think in practice it’s possible to identify a model as one of some specific family such as “Bayes Net”, “Neural Network”, “MaxEnt”, etc.
Perhaps a particular Bayes net together with a certain choice of conditional probability function for each arc and a certain choice of inference algorithm would constitute a model.
Right, the point is that the challenger can make any reasonable choice for these unspecified components. Ideally someone would say: here is the data set; I’m modeling it using the method described in such-and-such paper; here are some minor revisions to the method of the paper to make it useful in this case; here are the results.
Valid point, but I think in practice it’s possible to identify a model as one of some specific family such as “Bayes Net”, “Neural Network”, “MaxEnt”, etc.
Right, the point is that the challenger can make any reasonable choice for these unspecified components. Ideally someone would say: here is the data set; I’m modeling it using the method described in such-and-such paper; here are some minor revisions to the method of the paper to make it useful in this case; here are the results.