Search found 271 matches

by marek [BayesFusion]
Tue Jul 16, 2019 9:32 am
Forum: GeNIe
Topic: Intuitive noisy averager
Replies: 4
Views: 217

Re: Intuitive noisy averager

Cool, thanks for sharing this model with other GeNIe users. One suggestion, which is general, not necessarily about your model, as you may have other motivation for this. In the nodes "Net profit" and "Long-term biz opportunity" you have states that have zero a-priori probability. When they have zer...
by marek [BayesFusion]
Wed Jul 10, 2019 3:53 pm
Forum: GeNIe
Topic: Knowledge engineering with GeNIe
Replies: 5
Views: 232

Re: Knowledge engineering with GeNIe

Hi Charlie, I agree with you about the need to be patient when introducing this technology -- I have been teaching it for over 25 years :-). It is not easy at first but people love it once they understand it and it solves problems for them. By the way, BayesFusion offers free 30-day evaluation of Ba...
by marek [BayesFusion]
Tue Jul 09, 2019 10:28 am
Forum: GeNIe
Topic: Knowledge engineering with GeNIe
Replies: 5
Views: 232

Re: Knowledge engineering with GeNIe

Hi Charlie, Thanks again for sharing your experiences! In early days, Bayesian networks were almost exclusively based on knowledge engineering rather than being learned from data. Even though learning is powerful, it is precisely this subjective aspect of Bayesian network modeling and the ability to...
by marek [BayesFusion]
Tue Jul 09, 2019 9:48 am
Forum: GeNIe
Topic: Intuitive noisy averager
Replies: 4
Views: 217

Re: Intuitive noisy averager

Hi Charlie, Thanks for sharing this with the community. I have not used the Noisy-Adder/Average model in practice, so I am really pleased to see that it has useful applications. The attached model seems intuitive. I believe there is no difference between the two disconnected networks, correct? The N...
by marek [BayesFusion]
Mon Jun 24, 2019 9:47 pm
Forum: GeNIe
Topic: penalize parameter for accuracy
Replies: 4
Views: 181

Re: penalize parameter for accuracy

I have attempted to answer your question in a different thread. Now that you know how to change the decision threshold, you are influence the sensitivity of your minority class (the one that is most important for you). Please otherwise look at the general machine learning literature on unbalanced da...
by marek [BayesFusion]
Mon Jun 24, 2019 9:43 pm
Forum: GeNIe
Topic: unbalanced data and parameter learning
Replies: 2
Views: 108

Re: unbalanced data and parameter learning

There is no direct support for unbalanced classes in GeNIe, so you will have to use general machine learning results for unbalanced classes and prepare your training data set accordingly. If you prepare a balanced data set to learn parameters, please keep in mind that your prior probabilities of cla...
by marek [BayesFusion]
Tue Jun 18, 2019 10:19 am
Forum: GeNIe
Topic: penalize parameter for accuracy
Replies: 4
Views: 181

Re: penalize parameter for accuracy

When validating, GeNIe produces ROC curves. You can figure out from the ROC curve where you want to be (try hovering over it). Unfortunately, GeNIe does not allow you to change the decision/classification threshold (I believe this is what you mean). You can achieve what you want by producing an outp...
by marek [BayesFusion]
Wed Apr 24, 2019 12:28 pm
Forum: SMILE
Topic: EM algorithm and smoothing
Replies: 9
Views: 640

Re: EM algorithm and smoothing

I do not really know how to set the prior distribution of my variable (at least without doing this manually) I am confused by your question. How else would you like to set the prior distribution over your variable? ESS allows you to weigh the prior distribution with the data but you need a prior di...
by marek [BayesFusion]
Thu Mar 28, 2019 1:12 pm
Forum: GeNIe
Topic: Virtual evidence
Replies: 4
Views: 479

Re: Virtual evidence

I have examined the model. Everything is working fine, as expected. Virtual evidence does not set the node in question to a specified posterior probability distribution but rather offers uncertain evidence (specified by the virtual evidence distribution) for its states. The marginals are a function ...
by marek [BayesFusion]
Mon Mar 25, 2019 4:46 pm
Forum: GeNIe
Topic: Virtual evidence
Replies: 4
Views: 479

Re: Virtual evidence

Why setting virtual evidence for a node that has child nodes has no effect? If the node has no child node everything is all right.
I think it absolutely does have an effect in both cases. Would you perhaps attach an example where you think it should work but does not?
Cheers,

Marek
by marek [BayesFusion]
Mon Mar 25, 2019 4:43 pm
Forum: GeNIe
Topic: More info for noise adder please
Replies: 4
Views: 512

Re: More info for noise adder please

The Noisy-Adder model is described in the doctoral dissertation of Adam Zagorecki (2010), Section 5.3.1 Non-decomposable Noisy-average. Here is where you should be able to find it:

http://d-scholarship.pitt.edu/6542/

I hope this helps,

Marek
by marek [BayesFusion]
Mon Mar 18, 2019 8:44 pm
Forum: GeNIe
Topic: Submodel - a question + a suggestion
Replies: 4
Views: 475

Re: Submodel - a question + a suggestion

Hi Charlie. This is a conceptual GUI problem that we have been aware of from the very beginning. GeNIe includes several operations that may help. First of all, you can view any node's parents and children outside of the current sub-model by right-clicking on the small triangles on the left (parents)...
by marek [BayesFusion]
Mon Feb 25, 2019 5:52 pm
Forum: GeNIe
Topic: Any idea to create a truncated distribution?
Replies: 7
Views: 712

Re: Any idea to create a truncated distribution?

why GeNIe samples outside the bound anyway in the first place? Is this a waste of resources let along causing problems? This is a theoretical problem that is hard to solve in general. There are no general purpose sampling algorithms from truncated distributions, so each time that you sample from a ...
by marek [BayesFusion]
Mon Feb 25, 2019 5:43 pm
Forum: GeNIe
Topic: To make a truncated distribution
Replies: 5
Views: 625

Re: To make a truncated distribution

I now realized a problem associated with checking "reject out-of-bound and invalid sample" in Network Property - in a hybrid model, a child discrete chance node (as the "test" node in the attached example) won't get the probabilities of all of its states added to 100%. In the attached example, it a...
by marek [BayesFusion]
Tue Feb 05, 2019 3:29 pm
Forum: SMILE
Topic: Modelling ongoing development with Dynamic Bayesian Networks
Replies: 3
Views: 1111

Re: Modelling ongoing development with Dynamic Bayesian Networks

This does not happens in PYSMILE, as evidence in t+n can affect beliefs in t, but it is closer to what I will need to do later: restore the net from the DB, initialize with previous status (ANCHOR), add evidences (PLATE), gather results (TERMINAL), and store the results in the DB. I'm sorry I did n...