### Recent Posts

Pages: 1 ... 8 9 [10]
91
##### General Discussion / Discrete Boolean nodes with beliefs as data distributions?
« Last post by Marko Gerbec on January 05, 2015, 17:04:32  »
As a novice Hugin user and oriented to risk modelling applications, I am interested in using Boolean variables/nodes, however, the point values specified for parents for true and false states seem dull. Is it possible somehow to use "simple" distribution data (e.g, log-normal) for true state(s) and rest to 1 is set to false state?
Had read the Expressions chapter in Manual, and Distrubution () function as permitted for Boolean nodes is not clear enough for me.
92
##### HUGIN GUI Discussion / Re: Aligning Photos and cropping.
« Last post by Anders L Madsen on January 05, 2015, 10:23:01  »
93
##### General Discussion / Re: Expressions
« Last post by Frank Jensen on December 18, 2014, 15:25:30  »
I don't think there is an elegant way to do this.

However, if we assume that "f(y|z)" can be expressed using Hugin expression syntax, and we know that Z assumes one of a finite number of states, say z1, z2, and z3, then you can write max_z{f(y|z)} as follows:

max (f(y|z1), f(y|z2), f(y|z3))

Clearly not pretty, but it should work.

Frank
94
##### General Discussion / Re: Expressions
« Last post by Therese on December 17, 2014, 13:02:27  »
Here is an example:

I have a discrete binary node X, and an observation y which has a continuous distribution with density f(y | z) given a discrete network variable Z. I want to define the distribution of X | Z using the distribution of y. We may think of y as a fixed quantity that is used when building the network.

If I simply want

P(X = 1 | Z = z) = P(Y < y | Z = z),

I think that this could perhaps be done by instead making X an interval node with states [-inf, y] and [y, inf].

Now, sometimes I want to let

P(X = 1 | Z = z) = f(y | z)/k

and then P(X = 0 | Z = z) = 1 - P(X = 1 | Z = z).

This is more difficult for two reasons.

The first reason is the essence of my question: I need to be able to evaluate the density f in this arbitrary point y.

The other reason is that I need to identify a constant k so that f(y | z)/k becomes a well defined probability. When I create tables by hand, I may do this simply by taking k to be max_z{f(y | z)}. I suspect that this will not be possible, because the expression is naturally defined for only one state z of Z.

I hope that my question makes sense. Hugin already computes everything I need like a charm. I'm simply exploring other ways of doing the same thing

Therese
95
##### General Discussion / Re: Expressions
« Last post by Frank Jensen on December 16, 2014, 23:01:06  »
Hi Therese,

You can use Normal, LogNormal, and Gamma in expressions, see section 6.7.1 in the Hugin API Reference Manual.

But I suspect that it is not as simple as that.  Can you elaborate on the "slightly complicated to compute" part of your post?

Regards,
Frank
96
##### General Discussion / Expressions
« Last post by Therese on December 16, 2014, 19:10:30  »
Hi,

I have some rather large probability tables, where the probabilities are slightly complicated to compute in that they involve evaluating the pdf and cdf of a standard (continuous) family of distributions, e.g. the normal, lognormal, and gamma distributions.

Would it be possible for me to specify this by expressions rather than manually specifying the tables?
I suspect that perhaps not judging from the the set of building blocks for expressions that are given in C API and the GUI help pages, but I may have overlooked something.

All the best,
Therese
97
##### HUGIN GUI Discussion / Aligning Photos and cropping.
« Last post by Andrew Roberts on December 08, 2014, 20:20:42  »
Hello,

I am working on creating a comparison between many different camera settings. For this comparison I shot an object about 60 times as I cycled through the settings. Although I used a tripod, there is still a slight shift between photos.

I would like the align these photos and the crop them so only the object is showing. Therefore all the images would look identical except for the camera setting changes.

I am optimistic that this software is capable of the output I am looking for. However, I am new to the software and any help is appreciated!

Thanks!!
-Andrew
98
##### HUGIN GUI Discussion / Re: Default heap size
« Last post by Martin on September 29, 2014, 09:23:47  »
The default value used by the launcher is -Xmx512m.

The only way to set such options manually is starting hugin using the command line option as described in the other post.

Kind regards
Martin
99
##### HUGIN GUI Discussion / Default heap size
« Last post by mattdash on September 24, 2014, 16:29:35  »
Hello,

I'm using the command line option described here to specify the max heap size:
http://forum.hugin.com/index.php/topic,119.0.html

I'm wondering what default heap size is used when you run hugin.exe? And is there a way to change it from within the GUI, or only using the command line option described in the post linked above?

Thanks.
100
##### FAQ / Re: Analysis Wizard
« Last post by Anders L Madsen on September 08, 2014, 09:09:37  »
Hi,

Quote
In the Analysis Wizard, the Error rate for Test Data Accuracy pane of how much is acceptable?

This depends on the domain and the application. It may also be relevant to take a look at AUC of the ROC (which should be at least 0.5). By searching the internet you will be able to find rules of thumb on how to interpret/classify the performance of a classifier based on the AUC of the ROC.

Quote
In section Case table, The probability values 0.5 or less, Large difference between the actual data and test data are available (By multiplying the probabilities of each category for discrete data.), the cause is?what to do to fix it?

I do not understand this comment. If the performance of the model is not sufficient, then the model should be improved. When building a Bayesian network classifier from data a number of design choices have to be made, e.g., which variables to include, how do discretize numerical variables and which edges should be present in the model. It is impossible to say how you model could be improved without a detailed description of the model and data.
Pages: 1 ... 8 9 [10]