Free Trial

Safari Books Online is a digital library providing on-demand subscription access to thousands of learning resources.

Share this Page URL

Chapter 5. Decision Analysis Fundamentals > 5.2 Influence Diagrams - Pg. 195

5.2. INFLUENCE DIAGRAMS 195 5.2 Influence Diagrams In Section 5.1, we noted two difficulties with decision trees. First, the repre- sentation of a problem instance by a decision tree grows exponentially with the size of the instance. Second, the probabilities needed in a decision tree are not always the ones that are readily available to us. Next, we present an alternative representation of decision problem instances, namely influence diagrams, which do not have either of these difficulties. First, we only discuss representing prob- lem instances with influence diagrams. Then in Section 5.2.2 we discuss solving influence diagrams. 5.2.1 Representing with Influence Diagrams An i n f l u e n c e d i a g r a m contains three kinds of nodes: c h a n c e (or u n c e r - t a i n t y ) n o d e s representing random variables; d e c i s i o n n o d e s representing decisions to be made; and one u t i l i t y n o d e , which is a random variable whose possible values are the utilities of the outcomes. We depict these nodes as follows: -chance node