dynamic bayesian network inference

[safaa]

hello ;
1. what are the inference algorithm used in Dynamic Bayesian Network?
2. when we finish Dyanmic bayesian network and parameters are ready:how to extract the rules(predictions)?

[shooltz[BayesFusion]]

1. what are the inference algorithm used in Dynamic Bayesian Network?

The same you have selected for non-dynamic networks. We perform inference by transforming a DBN into unrolled BN, then run the algorithm and copy the posteriors from each timeslice back into the DBN.

2. when we finish Dyanmic bayesian network and parameters are ready:how to extract the rules(predictions)?

I'm not sure what you mean here.

[safaa]

2. when we finish Dyanmic bayesian network and parameters are ready:how to extract the rules(predictions)?

according to what I learned from the tutorials. to know the values of variables
1.choose value for one node
2. see what other values for each node will be
for example :a,b,c nodes
a=x then b =y in probability=z and c=m in probability=r
1. for all values of a, what are values of b and c
2. for all values of a in t , what are values of b in t+1 and c in t+1

[shooltz[BayesFusion]]

1.choose value for one node
2. see what other values for each node will be

You're not limited to 'one node'.
1. set the evidence (known values for zero, one or more nodes). Note that plate nodes will show the evidence UI which allows for time-specific evidence
2. update the network (if it's not set to update automatically in the Network menu)
3. inspect the posterior probabilities in one of the following ways:
a. show the 'Value' page of node properties
b. hover the mouse cursor over the checkmark icon displayed next to the node
c. enable the 'Temporal Beliefs' window from the 'View' menu and select the nodes/outcomes to display over time

[safaa]

thank you