Strength of influence

[mic]

Hi. Where can I find info about the features of the distance measures included in the strength of influence tool (Euclidean, Hellinger etc)?
How is the number appearing into the comment box calculated?
thank you!

[shooltz[BayesFusion]]

Here's the MSc thesis which describes the implementation of 'Strength of influence' feature in GeNIe:

http://www.kbs.twi.tudelft.nl/Publicati ... r-Msc.html

[mic]

very helpful! thank you

[mc]

Here's the MSc thesis which describes the implementation of 'Strength of influence' feature in GeNIe:

http://www.kbs.twi.tudelft.nl/Publicati ... r-Msc.html

I read the MSc thesis and I find it very useful! I am interested in better understanding of the "color of arcs" tool to show the influence sign. I tried to apply the tool with my network but it does not work (it is "not calculated"). Therefore I tried to built again one of the two thesis examples but they are not complete (not all the CPTs are showed).
Is it possible to have a functioning example?
Thank you!

PS: I found the problem! I had not set the ordinality of the nodes in the property dialog boxes of each node of my network. Now it works.
Anyway I think that is important to highlight the necessity of this setting to achieve the sign of influence analysis!

[mc]

I try to analyse my network through the strength of influence tool also applying the "color of arcs" in two different cases.
The first one is the network as I created it (let say the "natural" network) and the other one is with the introduction of decision nodes to make the sensitivity analysis (as you suggested in Tutorial 12). In this second network I settled as nominal all the decision node to obtain the same CPTs condition as in the "natural" network, but I could clearly find differences in the "color of arcs" analysis in the static apprach. Those differences disappear in the dynamic approach.
This is right?
Could someone help me?

Thank you very much.

mc