BayesFusion Support Forum

The answer to your question is unfortunately negative. It is not easy to pinpoint automatically which part of the model was responsible for misdiagnosis. This task can be best done manually.
Cheers,

Marek

Hi Elise2992, TAN learning algorithm tries to fit the joint probability distribution (jpd) modeled by the BN to the jpd that has generated the data. They make no attempt to discover the causal structure of the system under study. On the other hand, they are excellent when the data set is on the smal...

We have made modifications to the user interface between GeNIe 3.0 and 4.0 and removed this tab from node properties. The functionality is still available through the spreadsheet interface (columns on the right hand side of the spreadsheet). Our interactions with users of the diagnostic extensions i...

Thanks for the reminder. We are working on this and, unforeseen problems spared, should have it implemented roughly by mid-March (2023). We will share a stable prototype with you once it is ready.

Hi Elise, The color of the arc will be displayed once you specify for the nodes in your model whether they are increasing (Low to high) or decreasing (High to low) order of outcomes. I assume you have done that. The color green means that the higher the outcomes of the parent, the higher the outcome...

Hi Elise2992, Welcome to GeNIe/BN community! Let me try to answer your questions quickly: > 1. When making a BN using the program and wanting to do a validation > (K-fold, leave-one-out) afterwards do I need to divide my dataset in > a training and test set before making a BN? You can but you don't ...

Hi Paolo, What you would like to be able to model is asymmetry in the conditional probability distribution, something that is quite naturally modeled in probability trees. Unfortunately, this is not very natural for graphical models, like Bayesian networks. There was some work on this in early 1990s...

Perhaps one of GeNIe users will share their experience with multiple experts but I just wanted to point out that the question is really general and there are tons of answers. I have just performed a Google Scholar (https://scholar.google.com/) search with the following four terms: multiple experts B...

Hi Andy, You are right -- this is the way GeNIe/SMILE generate the samples when deriving the CPTs in the discretized version of the nodes. I believe that there is a brief description of this process in GeNIe manual. Please also note that in my DBN model, the variables are numerical, even though they...

Hi Andy, I'm afraid the link between DBNs and hybrid networks is still on our TO DO list. It is not something that will cause any serious theoretical problems, it is just that we haven't gotten to it. At the moment, unfortunately you will still need to use discrete variables. Doing things directly i...

You are completely right. We have screwed up the interface years ago while at the University of Pittsburgh by making the underlying implementation transparent at the GUI. It has been on our TO DO list to create a new canonical node, Noisy-AND/MIN that will simply be an AND/MIN, without making the us...

Hi Claire, You are probably referring to "soft evidence," which we don't have in GeNIe. We do have "virtual evidence" but this does not fix the posteriors, just offers uncertain evidence for them. You can read more about "virtual evidence' in GeNIe manual. Please do think ca...

Hi Andy, This is possible through cases. If you create or open a case library, you will be able to save the current states of evidence to a case. Cases are saved with the model, so you can load the case back once you have opened the model. You can find more information in GeNIe manual. Cheers, Marek

Hi Andy, I'm afraid you cannot do what you want automatically through GeNIe. The only way to achieve what you want is through SMILE, which is a general purpose library that can be used from inside your computer programs. Doing what you want should not be too hard using SMILE. I hope this helps, Marek

I'm afraid you will not be able to stop inference in a DBN in the middle, as we covert the DBN to a BN by unrolling and then run belief updating. You can control your inference by running it for a smaller number of steps, checking whether the condition that you were looking for has happened, and the...