Logic and the Foundations of Game and Decision Theory (LOFT 7)

Logic and the Foundations of Game and Decision Theory (LOFT 7)

GIACOMO BONANNO
WIEBE VAN DER HOEK
MICHAEL WOOLDRIDGE
Copyright Date: 2008
Pages: 240
https://www.jstor.org/stable/j.ctt46mz4h
  • Cite this Item
  • Book Info
    Logic and the Foundations of Game and Decision Theory (LOFT 7)
    Book Description:

    This volume collects together revised papers originally presented at the 7th Conference on Logic and the Foundations of Game and Decision Theory (LOFT 2006). LOFT is a key venue for presenting research at the intersection of logic, economics and computer science, and the present collection gives a lively and wide-ranging view of an exciting and rapidly growing area. This title is available in the OAPEN Library - http://www.oapen.org.

    eISBN: 978-90-485-0234-9
    Subjects: Economics, Philosophy, Mathematics

Table of Contents

  1. Front Matter
    (pp. 1-4)
  2. Table of Contents
    (pp. 5-6)
  3. Preface
    (pp. 7-10)
    G.B., W.v.d.H. and M.W.
  4. A Qualitative Theory of Dynamic Interactive Belief Revision
    (pp. 11-58)
    Alexandru Baltag and Sonja Smets

    This paper contributes to the recent and on-going work in the logical community [2, 14, 24, 8, 10, 9, 7] on dealing with mechanisms for belief revision and update within the Dynamic-Epistemic Logic (DEL) paradigm. DEL originates in the work of Gerbrandy and Groeneveld [30, 29], anticipated by Plaza in [44], and further developed by numerous authors [6, 31, 22, 4, 23, 39, 5, 15, 16] etc. In its standard incarnation, as presented e.g., in the recent textbook [25], the DEL approach is particularly well fit to deal withcomplex multi-agent learning actionsby which groups of interactive agents update...

  5. A Syntactic Approach to Rationality in Games with Ordinal Payoffs
    (pp. 59-86)
    Giacomo Bonanno

    The notion of rationalizability in games was introduced independently by Bernheim [2] and Pearce [16]. A strategy of playeriis said to be rational if it maximizes playeri’s expected payoff, given her probabilistic beliefs about the strategies used by her opponents; that is, if it can be justified by some beliefs about her opponents’ strategies. If playeri, besides being rational, also attributes rationality to her opponents, then she must only consider as possible strategies of her opponents that are themselves justifiable. If, furthermore, playeribelieves that her opponents believe that she is rational, then she must...

  6. Semantic Results for Ontic and Epistemic Change
    (pp. 87-118)
    Hans van Ditmarsch and Barteld Kooi

    In dynamic epistemic logics [32, 23, 9, 4, 17] one does not merely describe the static (knowledge and) beliefs of agents but also dynamic features: how does belief change as a result of events taking place. The main focus of such logics has been change ofonlybelief, whereas the facts describing the world remain the same. Change of belief is known asepistemicchange. One can also model change of facts, and the resulting consequences of such factual changes for the beliefs of the agents. Change of facts is also known asonticchange (change of the real world,...

  7. Social Laws and Anti-Social Behaviour
    (pp. 119-152)
    Wiebe van der Hoek, Mark Roberts and Michael Wooldridge

    A multiagent system [15] on the one hand aims at treating the agents as autonomous entities, whose behaviour should not be over-specified or too constrained, while at the same time one wants this system to achieve, or maintain certain objectives. As a consequence, one of the defining problems in multiagent systems research is that ofcoordination—managing the interdependencies between the actions of multiple interacting agents [1, 15]. There are broadly two techniques to approach this. Online techniques aim to equip agents with the ability to dynamically coordinate their activities, for example by explicitly reasoning about coordination at run-time. In...

  8. A Method for Reasoning about Other Agents’ Beliefs from Observations
    (pp. 153-182)
    Alexander Nittka and Richard Booth

    One of the overall goals of AI research is designing autonomous intelligent agents that are capable of acting successfully in dynamic environments. These environments may be artificial or even natural. In any case, it is very likely that they are “inhabited” by more than one agent. So, an agent will in general have to interact with (some of) the others. On the one hand, the agent—if it does not want to be purely reactive—needs a model of its environment in order to make informed choices of actions that change it in a way that brings the agent closer...

  9. A Logical Structure for Strategies
    (pp. 183-208)
    R. Ramanujam and Sunil Simon

    Extensive form turn-based games are trees whose nodes are game positions and branches represent moves of players. With each node is associated a player whose turn it is to move at that game position. A player’sstrategyis then simply a subtree which contains a unique successor for every node where it is this player’s turn to make a move, and contains all successors (from the game tree) for nodes where other players make moves. Thus a strategy is an advice function that tells a player what move to play when the game reaches any specific position. In two-player win/loss...

  10. Models of Awareness
    (pp. 209-240)
    Giacomo Sillari

    Since its first formulations (cf. [21]), epistemic logic has been confronted with the problem of logical omniscience. Although Kripkean semantics appeared to be the natural interpretation of logics meant to represent knowledge or belief, it implies that agents are reasoners that know (or at least are committed to knowing) every valid formula. Furthermore, agents’ knowledge is closed under logical consequence, so that if an agent knowsφandψis a logical consequence ofφ, then the agent knowsψas well. If we focus on representing pure knowledge attributions, rather than attributions of epistemic commitment, such a notion of...