Agent's Beliefs Research Articles

Subjective probability and quantum certainty

In the Bayesian approach to quantum mechanics, probabilities—and thus quantum states—represent an agent's degrees of belief, rather than corresponding to objective properties of physical systems. In this paper we investigate the concept of certainty in quantum mechanics. Particularly, we show how the probability-1 predictions derived from pure quantum states highlight a fundamental difference between our Bayesian approach, on the one hand, and Copenhagen and similar interpretations on the other. We first review the main arguments for the general claim that probabilities always represent degrees of belief. We then argue that a quantum state prepared by some physical device always depends on an agent's prior beliefs, implying that the probability-1 predictions derived from that state also depend on the agent's prior beliefs. Quantum certainty is therefore always some agent's certainty. Conversely, if facts about an experimental setup could imply agent-independent certainty for a measurement outcome, as in many Copenhagen-like interpretations, that outcome would effectively correspond to a preexisting system property. The idea that measurement outcomes occurring with certainty correspond to preexisting system properties is, however, in conflict with locality. We emphasize this by giving a version of an argument of Stairs [(1983). Quantum logic, realism, and value-definiteness. Philosophy of Science, 50, 578], which applies the Kochen–Specker theorem to an entangled bipartite system.

Ontology schema for an agent belief store

In this paper we explore the use of a formal ontology as a constraining framework for the belief store of a rational agent. The static beliefs of the agent are the axioms of the ontology. The dynamic beliefs are the descriptions of the individuals that are instances of the ontology classes. The individuals all have a unique identifier, an associated set of named classes to which they are believed to belong, and a set of property values. The ontology axioms act as a schema for the dynamic beliefs. Belief updates not conforming to the axioms lead to either rejection of the update or some other revision of the dynamic belief store to maintain consistency. Partial descriptions are augmented by inferences of property values and class memberships licensed by the axioms. For concreteness we sketch how such an ontology based agent belief store could be implemented in a multi-threaded logic programming language with action rules and object oriented programming features called Go!. This language was specifically designed for implementing communicating rational agent applications. We shall see that its logic rules allow us to extend an ontology of classes and properties with rule defined n-ary relations and functions. Its action rules enable us to implement a consistency maintenance system that takes into account justifications for beliefs. The pragmatics of consistency maintenance is an issue not normally considered by the ontology community. The paper assumes some familiarity with ontology specification using languages such as OWL DL and its subsets, and with logic programming.

Firm dynamics and markup variations: Implications for sunspot equilibria and endogenous economic fluctuations

This paper analyzes how the interaction between firms’ entry-and-exit decisions and variations in competition gives rise to self-fulfilling, expectation-driven fluctuations in aggregate economic activity and in measured total factor productivity (TFP). The analysis is based on a dynamic general equilibrium model in which net business formation is endogenously procyclical and leads to endogenous countercyclical variations in markups. This interaction leads to indeterminacy in which economic fluctuations occur as a result of self-fulfilling shifts in the beliefs of rational forward looking agents. When calibrated with empirically plausible parameter values and driven solely by self-fulfilling shocks to expectations, the model can quantitatively account for the main empirical regularities characterizing postwar U.S. business cycles and for 65 % of the fluctuations in measured TFP.

Departures from Rational Expectations and Asset Pricing Anomalies

We investigate the potential of the consumption CAPM with pessimism, doubt, and the availability heuristic in the agent's beliefs to resolve the equity premium and risk-free rate puzzles. Using the nonlinear GMM estimation techniques, we find that doubt and the availability heuristic play an important role in explaining the cross-section of asset returns. However, when taken alone, these deviations from rational expectations can not resolve the equity premium and risk-free rate puzzles. This result is robust to the assumption that the expected value of an uncertain prospect is nonlinear in the subjective outcome probabilities.

My beliefs about your beliefs: a case study in theory of mind and epistemic logic

We model three examples of beliefs that agents may have about other agents’ beliefs, and provide motivation for this conceptualization from the theory of mind literature. We assume a modal logical framework for modelling degrees of belief by partially ordered preference relations. In this setting, we describe that agents believe that other agents do not distinguish among their beliefs (‘no preferences’), that agents believe that the beliefs of other agents are in part as their own (‘my preferences’), and the special case that agents believe that the beliefs of other agents are exactly as their own (‘preference refinement’). This multi-agent belief interaction is frame characterizable. We provide examples for introspective agents. We investigate which of these forms of belief interaction are preserved under three common forms of belief revision.

Positive Self-Image and Incentives in Organizations

This paper investigates the implications of individuals' mistaken beliefs of their abilities on incentives in organizations using the principal-agent model of moral hazard. The paper shows that if effort is observable, then an agent's mistaken beliefs about own ability are always favorable to the principal. However, if effort is unobservable, then an agent's mistaken beliefs about own ability can be either favorable or unfavorable to the principal. The paper provides conditions under which an agent's over estimation about own ability is favorable to the principal when effort is unobservable. Finally, the paper shows that workers' mistaken beliefs about their coworkers' abilities make interdependent incentive schemes more attractive to firms than individualistic incentive schemes.

Monoparentalidade, pobreza e resiliência: entre as crenças dos profissionais e as possibilidades da convivência familiar

O presente artigo apresenta um estudo comparativo entre as crenças de agentes comunitários de saúde sobre possibilidades de resiliência em famílias monoparentais e de baixa renda e as estratégias de enfrentamento das adversidades que emergem nas histórias de vida destas famílias. Tomou-se como base a noção de resiliência aplicada à Psicologia, que se refere aos processos que explicam a superação de adversidades. Foram entrevistados quatro agentes comunitários de saúde e quatro famílias monoparentais. As estratégias metodológicas foram: entrevistas semi-estruturadas, histórias de vida, entrevista reflexiva e a análise dos dados através da grounded-theory. Os resultados revelaram crenças pessimistas dos agentes comunitários sobre o funcionamento das famílias pobres, em contraposição aos fatores indicativos de resiliência constatados nas histórias de vida destas famílias. Os resultados sinalizam para a necessidade de transformar e adequar as percepções dos agentes sociais acerca de grupos que vivem situação de pobreza.

Nursing Practice and Organisational Change Within the NHS: A Critical Realist Methodological Approach to the Analysis of Discursive Data

This methodological paper draws upon the results of a piece of qualitative research which explored nurses perceptions of the ways in which the organisational changes that occurred within the NHS over the past decade have mediated their working practices. This paper seeks to go beyond traditional forms of qualitative analysis in posing the question: Can the views and beliefs of social agents captured in social research inform a realist understanding of the social world? In addressing this question, the paper draws upon the insights of critical realist philosophy which takes as its central theoretical assumption, the interdependence of social structures and social interaction. This methodological approach conceptualises „discourse‟ (here defined as the articulation by social agents of their understanding of their world of social practices) as being as „real‟ as the structural interrelations to which it is materially linked through practice. Following Sayer (1997), analysing discourse in this way emphasizes its „performative‟ aspect rather than its representational aspect alone. Nevertheless, the accounts given by nurses as social agents are fallible and so interpretive analysis must involve a process of „theoretical transformation‟ or re-conceptualisation of these presentations of social practice. This paper sets out an analytical framework for the realist analysis of discourse which seeks to go beyond inductively or deductively-derived explanations of the social processes that shape nurses' working practices. Background The origins of this methodology paper lie in research conducted by the author, which was concerned with the impact of the organisational changes occurring within the NHS in the 1990‟s upon nursing practice. Nurses were then experiencing the consequences of new organisational priorities which were subjecting their practice to levels of managerial regulation and scrutiny not previously experienced by the profession. The working practices and attitudes of nurses were widely perceived by the Department of Health at the time to be old-fashioned and resistant to change, and therefore seen as representing an obstacle to the organisational goal of rationalisation and efficiency (DoH, 2002; Allen, 2001; Walsh and Gough 2000). The study itself sought to focus upon nurses‟ own perceptions of the ways in which these organisational changes mediated their working practices. The implicit aim of the research was to interrogate the nature of the interactive relationship that exists between structure and agency in the shaping of practice. This required firstly establishing what were the common (and deviant) themes that characterised the frames of meaning or „discourses‟ that these nurses drew upon in their day-to

Internal reasons and practical limits on rational deliberation

Could someone who wants a gin and tonic have a normative reason to drink petrol and tonic? Bernard Williams and Michael Smith both say, ‘No’. They argue that what an agent has normative reason to do is determined by rational deliberation that involves correcting the agent's beliefs and current motivations. On such an account of normative reasons, an agent who is motivated to act in some way due to a false belief does not have reason to act in that way. I argue that the agent could have reason to drink the petrol, because an agent's epistemic circumstances, what that agent can come to know, can be as relevant to what the agent has reason to do as other aspects of his circumstances. Moreover, if an agent's epistemic circumstances are taken into account when determining what the agent has reason to do, this can still give an account of reasons that is normative, ensures that the agent and onlookers agree on what the agent has reason to do, is appropriately connected to rationality and fairly represents the agent's beliefs about the knowledge they need to have to know how they have reason to act.

Dynamic Logic for Plan Revision in Agent Programming

In this paper, we present a dynamic logic for a propositional version of the agent programming language 3APL. A 3APL agent has beliefs and a plan. The execution of a plan changes an agent's beliefs. Plans can be revised during execution by means of plan revision rules. Due to these plan revision capabilities of 3APL agents, plans cannot be analyzed by structural induction as in for example standard propositional dynamic logic. We propose a dynamic logic that is tailored to handle the plan revision aspect of 3APL. The logic is one for plans that are restricted in a certain way. For this logic, we give a sound and complete axiomatization. Further, we discuss how this logic for restricted 3APL plans can be extended to a logic for non-restricted plans and we discuss some example proofs, using the logic. Finally, we consider the relation between proving properties of 3APL agents and proving properties of procedural programs.

Trust management and trust theory revision

A theory of trust for a given system consists of a set of rules that describe trust of agents in the system. In a certain logical framework, the theory is generally established based on the initial trust of agents in the security mechanisms of the system. Such a theory provides a foundation for reasoning about agent beliefs as well as security properties that the system may satisfy. However, trust changes dynamically. When agents lose their trust or gain new trust in a dynamic environment, the theory established based on the initial trust of agents in the system must be revised, otherwise it can no longer be used for any security purpose. This paper investigates the factors influencing trust of agents and discusses how to revise theories of trust in dynamic environments. A methodology for revising and managing theories of trust for multiagent systems is proposed. This methodology includes a method for modeling trust changes, a method for expressing theory changes, and a technique for obtaining a new theory based on a given trust change. The proposed approach is very general and can be applied to obtain an evolving theory of trust for agent-based systems

AGENT-ENHANCED COLLABORATIVE ACTIVITY IN ORGANIZED SETTINGS

For groups of agents to act collaboratively, they need to recognize the need for collaboration, decide on the method to be followed for achieving goal states, assign responsibilities to subgroups and individuals, and so on, until responsibilities that can be fulfilled by individuals are reached. Aiming to support collaborative activity of humans within organized settings, this paper introduces a set of constructs for specifying organizational structures and introduces an explicit representation of individual and collaborative responsibilities within organizations. We conjecture that group members create common awareness towards recognizing the need for collaboration by forming group acceptances. Acceptances are formed by means of shared practices and beliefs of individual agents. The paper introduces state recognition recipes that drive group members within organizations to form acceptances, and thoroughly explains the exploitation of these recipes in conjunction to state achievement recipes for achieving goal states and fulfilling responsibilities collaboratively.

Belief in Supernatural Agents in the Face of Death

Four studies examined whether awareness of mortality intensifies belief in supernatural agents among North Americans. In Studies 1 and 2, mortality salience led to more religiosity, stronger belief in God, and in divine intervention. In Studies 3 and 4, mortality salience increased supernatural agent beliefs even when supernatural agency was presented in a culturally alien context (divine Buddha in Study 3, Shamanic spirits in Study 4). The latter effects occurred primarily among the religiously affiliated, who were predominantly Christian. Implications for the role of supernatural agent beliefs in assuaging mortality concerns are discussed.

La politique de diffusion de l'information : la transparence des banques centrales est-elle toujours préférable ?

Communication Policy: is Central Bank Transparency Always Suitable? Transparency is one of the main pillars of central banks and international institutions. From their point of view, higher informational transparency improves their efficacy. The aim of this paper is to evaluate in what extent this strategy is beneficial or not. It proposes a survey of the literature based on coordination games. We show that an increase in informational transparency can lead to sub-optimal results. Indeed in providing a focal point for the beliefs of agents, public information facilitates coordination and so over-reaction behaviors. We do not deduce from this result that transparency is useless. We stress that it is necessary to adapt the communication policy to an environment characterized by strategic complementarities. To this end, we analyze several tools of information disclosure.

From Knowledge-based Programs to Graded Belief-based Programs, Part I: On-line Reasoning*

Knowledge-based programs (KBPs) are a powerful notion for expressing action policies in which branching conditions refer to implicit knowledge and call for a deliberation task at execution time. However, branching conditions in KBPs cannot refer to possibly erroneous beliefs or to graded belief, such as "if my belief that φ holds is high then do some action a else perform some sensing action β". The purpose of this paper is to build a framework where such programs can be expressed. In this paper we focus on the execution of such a program (a companion paper investigates issues relevant to the off-line evaluation and construction of such programs). We define a simple graded version of doxastic logic KD45 as the basis for the definition of belief-based programs. Then we study the way the agent's belief state is maintained when executing such programs, which calls for revising belief states by observations (possibly unreliable or imprecise) and progressing belief states by physical actions (which may have normal as well as exceptional effects).

Argumentation and the Dynamics of Warranted Beliefs in Changing Environments

One of the most difficult problems in Multi-Agent Systems (MAS) involves representing the knowledge and beliefs of an agent which performs its tasks in a dynamic environment. New perceptions modify this agent's current knowledge about the world, and consequently its beliefs about it also change. Such a revision and update process should be performed efficiently by the agent, particularly in the context of real-time constraints. In the last decade argumentation has evolved as a successful approach to formalize defeasible, commonsense reasoning, gaining wide acceptance in the MAS community by providing tools for designing and implementing features, which characterize reasoning capabilities in rational agents. In this paper we present a new argument-based formalism specifically designed for representing knowledge and beliefs of agents in dynamic environments, called Observation-based Defeasible Logic Programming (ODeLP). A simple but effective perception mechanism allows an ODeLP-based agent to model new incoming perceptions, and modify the agent's knowledge about the world accordingly. In addition, in order to improve the reactive capabilities of ODeLP-based agents, the process of computing beliefs in a changing environment is made computationally attractive by integrating a database with the agent's program, providing pre-compiled information about previous inferences. We present algorithms for managing dialectical databases as well as examples of their use in the context of real-world problems.

Perseus: Randomized Point-based Value Iteration for POMDPs

Partially observable Markov decision processes (POMDPs) form an attractive and principled framework for agent planning under uncertainty. Point-based approximate techniques for POMDPs compute a policy based on a finite set of points collected in advance from the agent's belief space. We present a randomized point-based value iteration algorithm called Perseus. The algorithm performs approximate value backup stages, ensuring that in each backup stage the value of each point in the belief set is improved; the key observation is that a single backup may improve the value of many belief points. Contrary to other point-based methods, Perseus backs up only a (randomly selected) subset of points in the belief set, sufficient for improving the value of each belief point in the set. We show how the same idea can be extended to dealing with continuous action spaces. Experimental results show the potential of Perseus in large scale POMDP problems.

WHY DOES IT MATTER THAT BELIEFS AND VALUATIONS BE CORRECTLY REPRESENTED?*

This article contains an analysis of a simple principal–agent problem illustrating possible problems that may arise when the prinicpal ascribes to the agent subjective probabilities and utilities that are implied by the subjective expected utility model but do not represent the agent's beliefs and valuations. In particular, it is possible that an incentive contract designed by the principal induces the agent to choose an action that is not in the principal's best interest.

A Framework for Sequential Planning in Multi-Agent Settings

This paper extends the framework of partially observable Markov decision processes (POMDPs) to multi-agent settings by incorporating the notion of agent models into the state space. Agents maintain beliefs over physical states of the environment and over models of other agents, and they use Bayesian updates to maintain their beliefs over time. The solutions map belief states to actions. Models of other agents may include their belief states and are related to agent types considered in games of incomplete information. We express the agents' autonomy by postulating that their models are not directly manipulable or observable by other agents. We show that important properties of POMDPs, such as convergence of value iteration, the rate of convergence, and piece-wise linearity and convexity of the value functions carry over to our framework. Our approach complements a more traditional approach to interactive settings which uses Nash equilibria as a solution paradigm. We seek to avoid some of the drawbacks of equilibria which may be non-unique and do not capture off-equilibrium behaviors. We do so at the cost of having to represent, process and continuously revise models of other agents. Since the agent's beliefs may be arbitrarily nested, the optimal solutions to decision making problems are only asymptotically computable. However, approximate belief updates and approximately optimal plans are computable. We illustrate our framework using a simple application domain, and we show examples of belief updates and value functions.

The Anti-Naturalism of Some Language Centered Accounts of Belief

Common sense explanations of human action are often framed in terms of an agent's beliefs and desires. Recent widely received views also take believing and desiring (as well wishing, intending and even thinking) as attitudes of an agent to linguistic or quasi-linguistic entities. It is here claimed that such a narrow view of cognitive attitudes is not supportable, since even among lingual non-verbal responses are often overriding evidence for belief and desire, even where they run counter to sincere verbal assents. The view is also curiously non naturalistic in that it disallows ascribing beliefs and desires altogether to non-lingual and pre-lingual. In the present paper a “common sense” explanation of action in accordance with the triad Desire, Belief, Action, is seen as a useful phenomenological “theory” provided that language centrality is not taken as essential.