AldenBlack
I understand, in principle, the difference the between searching with "Hard Evidence" and Value/Mean Variation with "mean" and "binary," but I don't see any difference in my results. More specifically, the solution using "hard evidence" is also a binary combination of two adjacent states (all my nodes are continuous), where I had expected the algorithm to return an answer that was 100% one of the states of the node (like usual hard evidence we would enter via the monitor panel). Is this the expected behavior, or did I configure something incorrectly?
Quote 0 0
Dan
This is clearly not the expected behavior. Can you post a screen shot of your Optimization settings?
Quote 0 0
AldenBlack
I'm having trouble getting the images to post properly - I'll send them to you in an email.
Quote 0 0