Recent Posts

1

HUGIN GUI Discussion / Re: Discrete vs Hybrid BN

« Last post by Anders L Madsen on January 19, 2023, 18:40:36  »

Hello,

Here are some quick replies to your questions

• It is easy to find VOI in the GUI however where and how I can see the posterior probability distribution for the target node?
  
  Each node has a monitor that displays the posterior probability distribution of the node. You need to compile the network, select the target node, right click on the target node and select toggle monitor window.
• The VOI is found in discrete but not in hybrid networks. So my question what are the common metrics that can be evaluated for both types of networks?
  
  Value of information in the GUI does not support hybrid variables.
  
• Also how are the sensitivity to findings and sensitivity to parameters evaluated in hybrid networks?
  
  In hybrid networks there is in many situations not a difference between a discrete observation and a continuous observation. Hence, the documentation applies to both situations. Parameter sensitivity analysis is not supported for the continuous parameters in hybrid networks
  

I hope this is useful
Anders

2

HUGIN GUI Discussion / Discrete vs Hybrid BN

« Last post by georgeskerr on January 16, 2023, 23:19:23  »

hello
I am trying to assess the performance of two models one is discrete and the other has the same structure but I kept continuous values.
My aim is to assess the sensitivity to findings " Value of information (VOI) " and sensitivity to parameters through the computation of the Posterior Probability Certainty index (through Posterior probability distribution), AUC, and ROC curves.

1.  It is easy to find VOI in the GUI however where and how I can see the posterior probability distribution for the target node?
2. The VOI is found in discrete but not in hybrid networks. So my question what are the common metrics that can be evaluated for both types of networks?
3. Also how are the sensitivity to findings and sensitivity to parameters evaluated in hybrid networks?

I would really appreciate your help. Also if someone has already done this comparison in performance for discrete and hybrid networks would really appreciate their suggestion and help.

3

General Discussion / A searched feature

« Last post by Piblo on July 08, 2022, 15:28:40  »

Hello,

Sorry if my post is in a wrong forum dscussion.

I'm looking for a feature that I can't find in Hugin. Either, it doesn't exist and it would be useful for many users, or I don't know how to find it.

I work on photography reconductions. I take pictures of views printed on old postcards... at the same location they were taken more than one century ago. I then mix them in a single photo.

Even with using equivalent focal lengths than for the originals, I must deform the recent picture in order it matches to the old one. As for panoramics with Hugin, I'm able to identify control points that would be used by a similar software. Until now, I manually modify the new picture thanks to GIMP and this is tedious and inaccurate.

What I would need, as output, is only the recent picture modified thanks to a process that use the control points... in order to use it later as a calc in GIMP. I don't wish the software automatically produces a mixed picture itself.

Please, could you help ? Thank you.

Regards,
Piblo   

4

General Discussion / Excel, Dates, and Discretization

« Last post by MarcoGV on May 03, 2022, 22:15:24  »

I recently ran into a problem that I do not know how to solve in a simple way.  It turns out that Excel automatically turns an entry like 5 - 10 into a date; in this case, May 10.  (I am in the U.S.)  (See https://support.microsoft.com/en-us/office/stop-automatically-changing-numbers-to-dates-452bd2db-cc96-47d1-81e4-72cec11c4ed8)  After discretization of a data set where one of the intervals is 5 - 10, the .cvs file built by Hugin has May 10 entries wherever 5 -10 should be.  Attempts to convert May 10 to 5 - 10 by formatting cells fail, because 5 - 10 is turned into a number representing the number of days from the "beginning of time," and trying to convert that number back to 5 - 10 results in May 10 again.  Does anyone have any suggestions to solve this problem?

5

HUGIN GUI Discussion / Re: A More General Zoom Feature

« Last post by MarcoGV on April 22, 2022, 17:28:14  »

The sizing of features in Hugin 9.1 is much better, at least in Windows.  Thank you for this improvement, which makes using Hugin more pleasant.

6

General Discussion / Re: cross-validation

« Last post by MarcoGV on April 21, 2022, 16:59:44  »

Are there any updates on the possibility of a cross-validation tool in Hugin?  I would find it very useful to complement the existing classifier performance tool.

7

General Discussion / Linking the interval values of the raster files with the nodes

« Last post by priyanka on March 07, 2022, 18:03:00  »

Hello Developers,

I am a beginner at Hugin software and trying to learn the software using HUGIN LITE. I wish to work with the raster files and for them, I am trying to build a Bayesian network. My problem is when I select the node type to interval and enter the values for the raster file (e.g. Indices, which has negative and positive values) manually in the table, it does not accept the values. I am not able to find any option where I can link the corresponding tiff file to the node. On youtube, the tutorial to build the model and attach or link the rasters to the node has not been explained but only to enter the values manually in the table of node has been demonstrated.

Please help me in this initial phase so that I can learn the model and further apply it in my research work.

Thanks in advance!

8

HUGIN GUI Discussion / Re: A More General Zoom Feature

« Last post by Anders L Madsen on September 04, 2021, 16:23:14  »

Marco,

We are working on a new type of monitors that will support better zoom functionality.

best regards
Anders

9

HUGIN GUI Discussion / Re: Information Entropy Minmization Discretization

« Last post by Anders L Madsen on September 04, 2021, 16:21:57  »

Hi Marco,

The IEM discretization method is based on the algorithm of the following paper:

U. M. Fayyad and K. B. Irani. 'Multi-Interval discretization of continuous valued attributes'. In Proceedings of the Thireenth International Joint Conference on Artificial Intelligence, pages 1022-1027, 1993.

If you have ideas for other (or better) discretization algorithms to implement, then please let us know.

Best regards
Anders

10

HUGIN GUI Discussion / Information Entropy Minmization Discretization

« Last post by MarcoGV on August 30, 2021, 20:18:13  »

Greetings!  Could someone please explain in more detail that in the manual what information entropy minimization discretization does? I tried it with a file with 10000 cases generated from the anginaContinuous network in Section 3.3.8 of Jensen and Nielsen's book, discretizing the Therm variable with the Angina variable as target.  It gave reasonable results, with three intervals.  Maybe someone could look at that example and use it to explain which entropy is minimized.