## Search found 1324 matches

Tue May 24, 2022 9:59 pm
Forum: SMILE
Topic: update cpt of a node regularly
Replies: 1
Views: 34

### Re: update cpt of a node regularly

To modify node's CPT, you should call set_node_definition. Refer to Tutorial 1 in SMILE Wrappers Manual - the tutorial program initializes the CPT with probabilities hardcoded into the Python source. In your case, you'll need to create an array from the CSV file, then call set_node_definition.
Fri Apr 15, 2022 3:51 pm
Forum: GeNIe
Replies: 1
Views: 250

### Re: explain the tornado diagram

most sensitivity parameter for "L = good" is "I = L | D = L". That means this paramter will make the posterior probability of "L = good" change from 0.303756 to 0.314423. The output probability range from 0.303756 to 0.314423 is calculated for parameter spread=10% (the...
Wed Apr 13, 2022 2:32 pm
Forum: SMILE
Topic: DLL load failed in python
Replies: 3
Views: 180

### Re: DLL load failed in python

The most likely cause for this error message would be 32-bit Python environment and 64-bit pysmile.pyd module.
Fri Apr 08, 2022 4:57 pm
Forum: GeNIe
Topic: node is not consistent
Replies: 4
Views: 178

### Re: node is not consistent

You can ignore the issue. Your model is closer to the perfect sum=1.0 distribution that our default epsilon.

Before we release new version, you can consider using the 'Complement' command in the columns with warnings to get perfect 1.0 total.
Wed Apr 06, 2022 12:57 pm
Forum: GeNIe
Topic: node is not consistent
Replies: 4
Views: 178

### Re: node is not consistent

We tracked down this problem to GeNIe's algorithm for epsilon selection. If a column in the CPT has a very small non-zero probability (like 2e-25 in your example), it will be used for comparison of the total probabilities in the column with 1.0. The default epsilon is on the order of 1e-6. This algo...
Thu Mar 31, 2022 9:43 pm
Forum: GeNIe
Topic: node is not consistent
Replies: 4
Views: 178

### Re: node is not consistent

I can't reproduce the problem. I have typed the numbers in your screenshots and got no warning icon with the 2e-25 value. Can you send me your xdsl file (use a forum private message if model contains confidential information). BTW: you can hover the mouse cursor over the warning icon sign and the pr...
Mon Mar 14, 2022 11:25 pm
Forum: GeNIe
Topic: problem with unrolling DBN
Replies: 4
Views: 179

### Re: problem with unrolling DBN

Unfortunately, GeNIe does not support equation nodes in DBNs.
Wed Mar 09, 2022 9:16 pm
Forum: GeNIe
Topic: how to interprete the result of sensitivity analysis
Replies: 4
Views: 232

### Re: how to interprete the result of sensitivity analysis

The tornado is ordered by the change in the target posterior when one of the parameters changes within the selected range. In your case the target is probability of the 'severe' outcome of the 'mental_health' node. The largest change in this probability would be caused by changing the appropriate en...
Tue Feb 08, 2022 1:36 pm
Forum: SMILE
Topic: All potential entries are zeros for node...
Replies: 1
Views: 374

### Re: All potential entries are zeros for node...

This error can be caused either by lost of arithmetic precision (for very large network), or by an inconsistent probabilities/evidence. Can you post your model so we can diagnose this? If it contains sensitive information, send me a private forum message with a download link.
Tue Feb 08, 2022 1:33 pm
Forum: GeNIe
Topic: Missing values in test set
Replies: 1
Views: 221

### Re: Missing values in test set

For each record in the test set we set evidence if the value is present in the record and the node is not a class node. The inference is then performed and most likely outcome is selected as prediction. Prediction is compared with actual value in the test set, and the confusion matrix is updated acc...
Tue Dec 14, 2021 9:57 pm
Forum: GeNIe
Topic: when use 17 millions data to learn the structure and parameter of network, it shows: abnormal program termination
Replies: 3
Views: 272

### Re: when use 17 millions data to learn the structure and parameter of network, it shows: abnormal program termination

Can you send me a link to your data file using private message on this forum?

Also, please note that GeNIe Academic is 32-bit only and its address space is limited to two gigabytes. I believe part of the problem here is dataset size.
Wed Oct 13, 2021 7:33 pm
Forum: SMILE
Topic: Maven repository for jSMILE is available now
Replies: 0
Views: 4590

### Maven repository for jSMILE is available now

For Maven-based Java builds it is now possible to use jSMILE from BayesFusion's Maven repository. For more details, read the "Platforms and Wrappers/Java and jSMILE/Maven" section in SMILE Wrappers manual at https://support.bayesfusion.com/docs.
Thu Aug 19, 2021 8:20 pm
Forum: GeNIe
Topic: EM calculations
Replies: 1
Views: 1439

### Re: EM calculations

Is there a way to obtain those?
The EM algorithm implementation in SMILE does not expose these values. They are discarded once EM's internal data structures are updated.
Thu Aug 05, 2021 12:14 pm
Forum: GeNIe
Topic: Crash DBN
Replies: 3
Views: 1729

### Re: Crash DBN

I found a cause of the problem; since I set one of the target variable as -1, it crashed.
Did you call SetEvidence and specified -1 as the outcome index?
Wed Jul 21, 2021 11:05 am
Forum: SMILE
Topic: Equation Nodes SMILE
Replies: 2
Views: 1812

### Re: Equation Nodes SMILE

You need to call set_cont_evidence (instead of set_evidence) on the continuous nodes. Please refer to SMILE Wrappers manual, section 5.12.2 titled "Continuous Inference". Below you can find the fixed code which I was able to run (I also fixed the problem with the format specifier in get_be...