BayesFusion Support Forum

Hi, I will be happy to help. When the Season is not observed, you get a mixture of Gaussians in the node Outside Air Temperature, weighted by the probabilities of the four seasons. I can see in the picture that you must have observed a value -2.5 in the node Outside Air Temperature (there is a groun...

I'm afraid that this may be tricky. Being able to refer to the posterior probabilities of the parent nodes would amount to using the solution of a system of equations in the definition of some of the variables. I would suggest that you make all parent nodes and the child node continuous and then ref...

SMILE unrolls a DBN and treats it as a static BN in which each of the time steps is represented explicitly. In this sense, given observations of any variables in any time steps, it calculates the conditional posterior probability distribution over the remaining variables. I believe that this procedu...

Hi sverrevr, It's been a while since I looked at the issue -- we implemented virtual evidence in GeNIe and SMILE quite possibly 20 years ago :-). The best source of information will be articles that contain the precise definition and formulas. Here are a couple that you may try: https://www.research...

You can do what you want by using the "Validate..." functionality in "Learning" menu. You can designate the variable that you are interested in to be the class node and also produce an output file. The output file contains the inputs and then additional columns for each of the st...

Hi, You can examine the machine diagnosis model interactively through our model repository. You can see each CPT when playing with the model. Here is the address of the repository, just in case you missed it: http://repo.bayesfusion.com/ The address provided to you by Tomek will lead you directly to...

Hi Lotte, I'd love to help but I'm not 100% sure what cutting value you are referring to. Do you mean the value of cross-entropy that makes an observation worthwhile? I'm quite sure that GeNIe does not support any such threshold directly. As far as value of information in general is concerned, the c...

Hi Danny, Please look at the attached network -- it implements comparison between two distances. Generally, you will have to use equation nodes for the comparison. If you insist on having a discrete network (there is really no need for that, as GeNIe implements hybrid networks with no limitations on...

Have you tried our network repository? You can download all models from the repository directly from inside GeNIe: Just try "Open from BayesBox..." The default BayesBox address is the address of our model repository. All these networks are, I believe, also installed on your disk drive (ins...

Hi Lotte, Somebody else at BayesFusion will answer your R question. Let me just comment on the entropy reduction question. The formula used in GeNIe/SMILE is somewhat different. We calculate cross-entropy, which is the expected difference between the entropy of the target and the entropy of the targ...

Hi Lotte, We support entropy-based value-of-information calculation, which is based on cross-entropy. Cross-entropy between two nodes T and E (Target and Evidence) is the expected reduction of entropy of T given that you will observe E. You can play with this in diagnostic extensions of GeNIe, QGeNI...

Hi Liid, The colors of nodes are based on the maximum value of sensitivity (i.e., highest for any parameter). The SA algorithm performs calculation for all nodes because it is very, very efficient and it does not makes practical sense to run it on a subset of nodes when you can calculate sensitiviti...

Hi Liid, You are right in that unrolling the network may be your only option. It is theoretically sound, as one way of conceptualizing inference in DBNs is that they are first unrolled. What will happen is that you may see the same patterns in various slices, depending of course what you select as t...

Hi Liid, Let me try to answer your questions one by one: (1) For learning DBN parameters, all your data should be in one file that follows the format described in GeNIe manual. In that file, you will have all time slices, i.e., ds0, ds1 and ds2. Otherwise, you are right that the ds0 data will be use...

Hi Camilla, I have always thought about this (i.e., modeling) as common sense, so nothing comes to mind in terms of the literature. As far as behavior of distributions with long tails goes, there is plenty of literature, including books -- a Google search will get you better results than what I can ...