06.03.2013 Views

Artificial Intelligence and Soft Computing: Behavioral ... - Arteimi.info

Artificial Intelligence and Soft Computing: Behavioral ... - Arteimi.info

Artificial Intelligence and Soft Computing: Behavioral ... - Arteimi.info

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

We now compute the messages that node V sends to its parents U <strong>and</strong> each of<br />

its children Z1, Z2, …,Zn to update their values. Each of these two messages is<br />

a conditional probability, given that the condition holds <strong>and</strong> the probability<br />

given that it does not.<br />

Now, the message from V to parent U, denoted by λv (U), is computed as<br />

λv(U) = ∑v ∈(true, false) P(ev - |U, V = v) P(V = v | U)<br />

= ∑v∈(true,false) P(ev - |,V= v) P (V = v |U)<br />

= ∑ v ∈(true,false) P(V = v |U) λ (V = v)<br />

= [P(V | U)]2 x 2 x [λ(0) λ(1)] T 2 x 1 (9.17)<br />

Lastly, the message from V to its child Zj is given by<br />

∏ Zj (V)<br />

= P(V⏐eZi + )<br />

= P(V⏐ev + , eZ1 - , eZ2 - , ……, eZi-i - , eZi+ i - ,………, eZn - )<br />

= β ∏j≠i P(eZi - ⏐V, ev + ) P(V⏐ev + )<br />

= β ∏j≠iP(ezi - ⏐V) P(V⏐ev + )<br />

=β( ∏j≠iλZi (V)) ∏(V)<br />

=β(λ(V)/ λZj(V)) ∏(V)<br />

=β Bel (V)/ λZj(V) (9.18)<br />

where β is a normalizing constant computed similarly as α.<br />

The belief updating process at a given node B (in fig. 9.4) has been<br />

illustrated based on the above expressions for computing the λ <strong>and</strong> ∏<br />

messages. We here assumed that at each node <strong>and</strong> link of the tree (fig. 9.4) we<br />

have one processor [7]. We call these node <strong>and</strong> link processor respectively. The<br />

functions of the node <strong>and</strong> the link processors are described in fig. 9.6.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!