11.7 Information Spaces in Game Theory

This section unifies the sequential game theory concepts from Section 10.5 with the I-space concepts from this chapter. Considerable attention is devoted to the modeling of information in game theory. The problem is complicated by the fact that each player has its own frame of reference, and hence its own I-space. Game solution concepts, such as saddle points or Nash equilibria, depend critically on the information available to each player as it makes it decisions. Paralleling Section 10.5, the current section first covers I-states in game trees, followed by I-states for games on state spaces. The presentation in this section will be confined to the case in which the state space and stages are finite. The formulation of I-spaces extends naturally to countably infinite or continuous state spaces, action spaces, and stages [59].


Steven M LaValle 2012-04-20