Genie NoisyAdder

The front end.
Post Reply
UNSWPHD
Posts: 2
Joined: Mon Jul 01, 2013 6:54 am

Genie NoisyAdder

Post by UNSWPHD »

I am using Genie within my PhD to model Political-Strategic Making. I have opted to use the PICI Noisy Average (Noisy Adder) model as a viable means to model cause effect relationships. However, I am particularlyinterested in situations where there are no significant non-modelled causes - i.e. no Leaks. I can't see how this is done in Genie as it seems some leak must always be present. This is complicated by a lack of documentation on Noisy Adders. I was hoping to replicate the results in the Adam Zagorecki's PhD example (reference is on the website) as a way to build confidence in how I am using the method. Can you please advise?
UNSWPHD
Posts: 2
Joined: Mon Jul 01, 2013 6:54 am

Re: Genie NoisyAdder

Post by UNSWPHD »

As I now have found that I can get the noisy adder to work by both setting the weight of the Leak variable to very low (< 10^-7) and ensuring that the leak probability distribution follows the distinguished state of the child, please consider this post closed.
Post Reply