Search found 430 matches
- Wed Sep 07, 2022 9:11 pm
- Forum: GeNIe
- Topic: Adaptative time step in DBN inference
- Replies: 4
- Views: 4004
Re: Adaptative time step in DBN inference
I'm afraid GeNIe will not do automatically what you want to do. Perhaps you can try running the model at a rougher time step to figure out where interesting things are happening and then with a finer time step in places that you discover in the first run? If you do it through SMILE, it should not be...
- Thu Sep 01, 2022 5:58 pm
- Forum: SMILE
- Topic: node evidence
- Replies: 3
- Views: 3881
Re: node evidence
I am a little confused, as in the Bayesian networks world we call the probability tables that are the definition of a node CPTs, Conditional Probability Tables. They are indexed not by evidence but by the various combinations of the states of parents of the node. It is possible that they mean someth...
- Mon Aug 29, 2022 3:30 pm
- Forum: SMILE
- Topic: node evidence
- Replies: 3
- Views: 3881
Re: node evidence
Hi Sara, "Evidence" is a very basic term in Bayesian inference and denotes an observation of a concrete state of a node. Your question is about SMILE programming and there are most certainly people at BayesFusion and outside who know more than I do about SMILE programming. You can get exam...
- Tue Aug 16, 2022 8:25 pm
- Forum: SMILE
- Topic: the representation of the bayesian network in a meta ontology
- Replies: 1
- Views: 3576
Re: the representation of the bayesian network in a meta ontology
Hi Sara, I would be happy to help but you are using vocabulary that I am not really familiar with. What do you mean by "meta model"? What do you mean by "instantiating" a Bayesian network? In the Bayesian network world, it usually means assigning values to an existing Bayesian ne...
- Tue Aug 09, 2022 7:33 pm
- Forum: GeNIe
- Topic: Promblem with mapping fault tree into bayesian network on GeNIe
- Replies: 1
- Views: 2829
Re: Promblem with mapping fault tree into bayesian network on GeNIe
Your error is assuming that E1 and E2 are independent. They are not, as they have a common ancestor, b. Because they are dependent, you cannot just multiply probabilities as you did in your calculation.
I hope this helps,
Marek
I hope this helps,
Marek
- Wed Jul 27, 2022 6:25 am
- Forum: SMILE
- Topic: PCTs updates
- Replies: 1
- Views: 3558
Re: PCTs updates
Absolutely! Start with "Learning parameters" from the "Learning" menu. In the "Learn Parameters with EM" dialog, select "Keep original" (parameters) and give the existing data a confidence level that corresponds to the number of records that the existing param...
- Wed Jul 13, 2022 10:37 am
- Forum: GeNIe
- Topic: Is there a way to instantiate general variables in probability tables of discrete nodes?
- Replies: 2
- Views: 3076
Re: Is there a way to instantiate general variables in probability tables of discrete nodes?
I'm afraid you cannot do this in GeNIe. The fastest way would be copying and pasting the entire table from the source to the destination nodes. This, I'm afraid, is not what you would like. It seems to me that you should program this in your favorite programming language and use SMILE. This will be ...
- Fri Jul 08, 2022 3:56 pm
- Forum: GeNIe
- Topic: Inquiries regarding sampling methods used by SMILE engine
- Replies: 2
- Views: 3045
Re: Inquiries regarding sampling methods used by SMILE engine
Hi Juan Luis, Thank you for your kind words about GeNIe and SMILE -- we have been trying our best to develop a flexible and useful tool for the academic and commercial communities. You will find the GIS capabilities included in the most recent release of GeNIe (4.0) very useful in environmental work...
- Thu Jun 30, 2022 9:09 pm
- Forum: GeNIe
- Topic: Too many states in nature nodes
- Replies: 3
- Views: 4030
Re: Too many states in nature nodes
I roughly see what you want to accomplish. I'm afraid your CPT in node PorductHierarchyID will have 40,000 conditional probability distributions and given that the node itself has 50 states, it will have 2,000,000 probabilities. If you multiply these by 8 bytes needed to represent a floating point n...
- Tue Jun 28, 2022 7:16 pm
- Forum: GeNIe
- Topic: Too many states in nature nodes
- Replies: 3
- Views: 4030
Re: Too many states in nature nodes
Generally, probabilistic representations are wonderfully efficient but only if there are independencies in the domain. Still, probability tables grow exponentially with the number of parents. The base of the exponent is the number of states in the node and parents. Having a node with thousands of st...
- Tue Jun 21, 2022 7:15 pm
- Forum: GeNIe
- Topic: randomize parameter
- Replies: 2
- Views: 2196
Re: randomize parameter
Hi Yan, I believe you are asking about randomization in the context of the EM algorithm (please correct me if I am wrong). In validation, this parameter also refers to the EM algorithm part (learning the parameters at each cross-validation step). It is an empirical finding that randomization gets yo...
- Thu Jun 16, 2022 11:39 am
- Forum: GeNIe
- Topic: Help with influence diagram
- Replies: 4
- Views: 3482
Re: Help with influence diagram
Clearly, your influence diagram is too complex. Please keep in mind that probabilistic inference is worst case NP-hard and it is not that difficult to construct models that are not computable. Quite likely if a model cannot be solved by SMILE/GeNIe, you cannot solve it with any software on Earth. Th...
- Tue Jun 14, 2022 3:08 pm
- Forum: GeNIe
- Topic: Help with influence diagram
- Replies: 4
- Views: 3482
Re: Help with influence diagram
Your model is quite impressive (in terms of size and complexity). I am not sure you can solve it using our default ID algorithm, as it will involve calculating the expected utilities over 2^19 (524,288) different policies. This is, of course, when you change back the RCM* nodes to decision nodes. Ar...
- Thu Jun 09, 2022 6:51 pm
- Forum: GeNIe
- Topic: query about noisyMAX algorithm
- Replies: 9
- Views: 3725
Re: query about noisyMAX algorithm
I see. I misunderstood your example but have have clarified it sufficiently. When A and B have an unobserved common parent C, you can write the following: P(A,B)=Sigma_over_i P(A|B,C_i)P(B|C_i)P(C_i) You write this formula for all possible states of C_i. Because both A and B are dependent on C, you ...
- Wed Jun 08, 2022 9:33 am
- Forum: GeNIe
- Topic: query about noisyMAX algorithm
- Replies: 9
- Views: 3725
Re: query about noisyMAX algorithm
Please look at a good Bayesian networks textbook. Pearl 1988 perhaps? When two adjacent variables are dependent, you can see it in the CPT: the distributions are different for different states of the parent. P(A,B)=P(A|B)P(B) for A and B being dependent. When they are independent, P(A|B)=P(A), i.e.,...