We propose a model

We describe a model of iterated belief revision that extends the AGM theory of revision to account for the effect of a revision on the conditional beliefs of an agent. In particular, this model ensures that an agent makes as few changes as possible to the conditional component of its belief set. Adopting the Ramsey test, minimal conditional revision provides acceptance conditions for arbitrary right-nested conditionals. We show that problem of determining acceptance of any such nested conditional can be reduced to acceptance tests for unnested conditionals. Thus, iterated revision can be accomplished in a “virtual” manner, using uniterated revision.

Belief revision and belief update have been proposed as two types of belief change serving different purposes, revision intended to capture changes in belief state reflecting new information about a static world, and update intended to capture changes of belief in response to a changing world. We argue that routine belief change involves elements of both and present a model of generalized update that allows updates in response to external changes to inform an agent about its prior beliefs. This model of update combines aspects of revision and update, providing a more realistic characterization of belief change. We show that, under certain assumptions, the original update postulates are satisfied. We also demonstrate that plain revision and plain update are special cases of our model. We also draw parallels to models of stochastic dynamical systems, and use this to develop a model that deals with iterated update and noisy observations in (qualitative settings) that is analogous to Bayesian updating in a quantitative setting. Some parts of this report appeared in preliminary form in “Generalized Update: Belief Change in Dynamic Settings,” Proc. of Fourteenth International Joint Conf. on Artificial Intelligence (IJCAI-95), Montreal, pp.1550–1556 (1995).

Research in belief revision has been dominated by work that lies firmly within the classic AGM paradigm, characterized by a well-known set of postulates governing the behavior of “rational” revision functions. A postulate that is rarely criticized is the success postulate: the result of revising by an observed proposition'results in belief in'. This postulate, however, is often undesirable in settings where an agent’s observations may be imprecise or noisy. We propose a semantics that captures a new ontology for studying revision functions, which can handle noisy observations in a natural way while retaining the classical AGM model as a special case. We present a characterization theorem for our semantics, and describe a number of natural specialcases that allow ease of specification and reasoning with revision functions. In particular, by making the Markov assumption, we can easily specify and reason about revision.

In this paper we describe two approaches to the revision of probability functions. We assume that a probabilistic state of belief is captured by a counterfactual probability or Popper function, the revision of which determines a new Popper function. We describe methods whereby the original function determines the nature of the revised function. The first is based on a probabilistic extension of Spohn's OCFs, while the second exploits the structure implicit in the Popper function itself. This stands in contrast with previous approaches that associate a unique Popper function with each absolute (classical) probability function. We also describe iterated revision using these models. Finally, we consider the point of view that Popper functions may be abstract representations of certain types of absolute probability functions, but show that our revision methods cannot be naturally interpreted as conditionalization on these functions. Keywords: belief revision, probability, Popper functions...