Properties of states of recongnition arising from a set of knowledge systems with imperfect reporting^†

† The Research of this author was supported in part by NASA Grant NGT 44-001-800.

Abstract

The main purpose of this article b to study how sometimes incomplete information coming from different sources can be put together. After interrogating the q^th source, a variable X_q is defined whose values are passible states. It is shown that {X_q } forms a Markov chain. It is shown that the probability of not getting additional information at the q^th source is the averaged commonality induced by X_q . Other related questions are looked at. Then, strategies are defined as to what to do when the probability of different states is not known. A possible way to proceed is to view possible states as environments in which decisions must be made. Different strategies, under different degrees of the reliability of the sources, are sketched. The belief function is used to build different payoffs and fuzzy versions of the problems are studied

† The Research of this author was supported in part by NASA Grant NGT 44-001-800.

† The Research of this author was supported in part by NASA Grant NGT 44-001-800.

Notes

† The Research of this author was supported in part by NASA Grant NGT 44-001-800.