next up previous
Next: Generating the diagram Up: Bayes linear influence diagrams Previous: Establishing and generating the

Assessing information content and flow for each yield separately

There are many different explorations of the beliefs and data that we could make. For our first example, we will assess the value of observing the collection tex2html_wrap_inline2001 and the extra value of observing the collection tex2html_wrap_inline2003 . We will draw influence diagrams to portray the expected effects of adjustment, calculated a priori, together with diagnostics comparing what we expected to what actually happened. We will examine individual aspects of the belief adjustment with respect to each of the four slope and intercept quantities tex2html_wrap_inline1945 , as well as overall aspects of the belief adjustment with respect to the collection tex2html_wrap_inline1963 containing them. To do this, it is necessary first to design the influence diagram, and second to carry out the relevant adjustments.



David Wooff
Thu Oct 15 11:27:04 BST 1998