A note on irrelevance in inductive logic (Q543531): Difference between revisions

From MaRDI portal
Added link to MaRDI item.
ReferenceBot (talk | contribs)
Changed an Item
 
(4 intermediate revisions by 3 users not shown)
Property / author
 
Property / author: Jeffrey Bruce Paris / rank
Normal rank
 
Property / author
 
Property / author: Alena Vencovská / rank
Normal rank
 
Property / author
 
Property / author: Jeffrey Bruce Paris / rank
 
Normal rank
Property / author
 
Property / author: Alena Vencovská / rank
 
Normal rank
Property / MaRDI profile type
 
Property / MaRDI profile type: MaRDI publication profile / rank
 
Normal rank
Property / full work available at URL
 
Property / full work available at URL: https://doi.org/10.1007/s10992-010-9154-2 / rank
 
Normal rank
Property / OpenAlex ID
 
Property / OpenAlex ID: W2132141480 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q5813549 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q5658081 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Concerning measures in first order calculi / rank
 
Normal rank
Property / cites work
 
Property / cites work: Some observations on induction in predicate probabilistic reasoning / rank
 
Normal rank
Property / cites work
 
Property / cites work: I.—PROBABILITY: THE DEDUCTIVE AND INDUCTIVE PROBLEMS / rank
 
Normal rank
Property / cites work
 
Property / cites work: A survey of some recent results on spectrum exchangeability in polyadic inductive logic / rank
 
Normal rank
Property / cites work
 
Property / cites work: Representation theorems for probability functions satisfying spectrum exchangeability in inductive logic / rank
 
Normal rank
Property / cites work
 
Property / cites work: Explication of inductive probability / rank
 
Normal rank
Property / cites work
 
Property / cites work: A continuum of inductive methods arising from a generalized principle of instantial relevance / rank
 
Normal rank
Property / cites work
 
Property / cites work: Common sense and maximum entropy / rank
 
Normal rank
Property / cites work
 
Property / cites work: The Uncertain Reasoner's Companion / rank
 
Normal rank
Property / cites work
 
Property / cites work: A note on the inevitability of maximum entropy / rank
 
Normal rank
Property / cites work
 
Property / cites work: Some aspects of polyadic inductive logic / rank
 
Normal rank
Property / cites work
 
Property / cites work: ON FILLING-IN MISSING CONDITIONAL PROBABILITIES IN CAUSAL NETWORKS / rank
 
Normal rank
Property / cites work
 
Property / cites work: Atom exchangeability and instantial relevance / rank
 
Normal rank

Latest revision as of 05:02, 4 July 2024

scientific article
Language Label Description Also known as
English
A note on irrelevance in inductive logic
scientific article

    Statements

    A note on irrelevance in inductive logic (English)
    0 references
    17 June 2011
    0 references
    This paper can be seen as a call to arms to better understand relevance and irrelevance in Carnap's framework of inductive (monadic) predicate logic. The context is the usual, i.e., there are countably many constants \(a_1,a_2\dots\) that are said to exhaust the universe, finitely many unary relation symbols \(P_1,\dots,P_q\), and the usual logical connectives \(\vee,\wedge,\neg,\forall,\exists.\) Let \textit{SL} and \textit{QFSL} be the sentences, respectively quantifier-free sentences, of this language. The authors motivate and formulate a notion of irrelevance every probability function \(w\) on \textit{QFSL} should satisfy. The notion is called Weak Irrelevance Principle (WIP), which states that, for all \(\varphi,\theta\in {\mathit QFSL}\) which do not mention a relation symbol or constant in common, it holds that \(w(\theta|\varphi)=w(\theta).\) The authors then introduce a family of probability functions \(u^{\overline p,L}_n\) for which it holds that \(\lim_{n\to\infty}u^{\overline p,L}_n=u^{\overline p,L}.\) The \(u^{\overline p,L}\) have also appeared in other papers by the authors and by the reviewer, where they were used to study polyadic inductive logic, in particular for the study of Language Invariance (LI) and Instantial Relevance (IR); cf. [\textit{J. Landes, J. Paris} and \textit{A. Vencovská}, Lect. Notes Comput. Sci. 5378, 162--169 (2009; Zbl 1203.03031)]. The authors then show that the \(u^{\overline p,L}_n\) and \(u^{\overline p,L}\) satisfy WIP and the principles of Constant Exchangeability (Ex) and Atom Exchangeability (Ax), furthermore the \(u^{\overline p,L}_n\) and \(u^{\overline p,L}\) satisfy LI with Ex, Ax and WIP. The next result presented is that the \(u^{\overline p,L}_n\) are the only probability functions satisfying LI with Ex, Ax and WIP. As a corollary it follows that the \(u^{\overline p,L}_n\) are in fact equal to some \(u^{\overline q,L}\). The \(u^{\overline q,L}\) are generalizations of the Paris-Nix continuum (given by the \(w^{\delta}\)). Thus, the \(w^{\delta}\) satisfy LI with Ex, Ax, WIP and the Generalized Principle of Instantial Relevance (GPIR). The intriguing part is that also the \(w^{\delta}\) appear to be an utterly reasonable choice for a rational agent choosing a probability function representing his beliefs, the \(w^{\delta}\) do not in general satisfy the classical Johnson Sufficientness Principle (JSP). Only the extreme points of continuum satisfy JSP, namely for \(w^0=c_{\infty}\) and \(w^{1}=c_0\), where the \(c_{\lambda}\) are the extreme points of Carnap's influential continuum.
    0 references
    0 references
    0 references
    0 references
    0 references
    uncertain reasoning
    0 references
    inductive logic
    0 references
    weak irrelevance principle
    0 references
    generalized principle of instantial relevance
    0 references
    Paris-Nix continuum
    0 references
    Johnson's sufficientness postulate
    0 references
    0 references
    0 references
    0 references