Conditional Utility, Utility Independence, and Utility Networks

We introduce a new interpretation of two re­ lated notions conditional utility and utility independence. Unlike the traditional inter­ pretation, the new interpretation render the notions the direct analogues of their prob­ abilistic counterparts. To capture these no­ tions formally, we appeal to the notion of util­ ity distribution, introduced in previous paper. We show that utility distributions, which have a structure that is identical to that of probability distributions, can be viewed as a special case of an additive multiattribute utility functions, and show how this special case permits us to capture the novel senses of conditional utility and utility independence. Finally, we present the notion of utility net­ works, which do for utilities what Bayesian networks do for probabilities. Specifically, utility networks exploit the new interpreta­ tion of conditional utility and utility indepen­ dence to compactly represent a utility distri­ bution.