Stochastic independence, causal independence, and shieldability

Spohn, Wolfgang

doi:10.1007/bf00258078

Cited by 143 publications

(70 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…(12) yields 2 The graphoid axioms are axioms of conditional independence, first formulated by Dawid [3] and Spohn [4]. Their connections to graph connectivity and to other notions of "information relevance" were established by Pearl and Paz [5] …”

Section: Deriving Testables From Non-testablesmentioning

confidence: 99%

“…Not less important is the ability of these axioms to justify ignorability relations which a researcher may need for deriving causal effect estimands [1,8,11,12]. 4 Consider the sentence Z x ? ?…”

Section: Deriving Ignorability Relationsmentioning

confidence: 99%

“…? Y z j X would allow 4 Reliance on the assumptions of conditional ignorability [8,11,12], which are cognitively formidable, is one of the major weaknesses of the potential outcome framework us to obtain the desired effect PðY z ¼ yÞ by adjusting for X, as shown in the derivation of eq. (25).…”

Section: Deriving Ignorability Relationsmentioning

confidence: 99%

See 2 more Smart Citations

Graphoids over Counterfactuals

Pearl

2014

Journal of Causal Inference

View full text Add to dashboard Cite

show abstract

Section: Deriving Testables From Non-testablesmentioning

confidence: 99%

Section: Deriving Ignorability Relationsmentioning

confidence: 99%

See 1 more Smart Citation

Graphoids over Counterfactuals

Pearl

2014

Journal of Causal Inference

View full text Add to dashboard Cite

show abstract

“…The axioms for semi-graphoids are presented in Pearl (1988, 82-90). They first occur in Dawid (1979) and Spohn (1980). For details on the d-separation criterion, see Pearl (1988, 117-118), Neapolitan (1990, 202-207) and Jensen (1996, 12-14).…”

Section: Modeling Confirmation With a Ltfr Instrumentmentioning

confidence: 99%

Bayesian Networks and the Problem of Unreliable Instruments

Bovens

Hartmann

2002

Philos. of Sci.

View full text Add to dashboard Cite

Original citation:Bovens, Luc and Hartmann, Stephan (2002) We appeal to the theory of Bayesian Networks to model different strategies for obtaining confirmation for a hypothesis from experimental test results provided by less than fully reliable instruments. In particular, we consider (i) repeated measurements of a single test consequence of the hypothesis, (ii) measurements of multiple test consequences of the hypothesis, (iii) theoretical support for the reliability of the instrument, and (iv) calibration procedures. We evaluate these strategies on their relative merits under idealized conditions and show some surprising repercussions on the variety-ofevidence thesis and the Duhem-Quine thesis.

show abstract

“…Th e intended interpretation of I(X, Y r Z ) (read X is independent of Y given Z ) is that having observed Z, no additional information about X could be obtained by also observing Y. For example, in a probabilistic model (Dawid 1979, Geiger et al 1991, Lauritzen et al 1990, Spohn 1980, Studeny! 1989…”

Section: Preliminariesmentioning

confidence: 99%

Independency relationships and learning algorithms for singly connected networks

Campos¹

1998

Journal of Experimental & Theoretical Artificial Intelligen

View full text Add to dashboard Cite

Abstract. Graphical structures such as Bayesian networks or M arkov networks are very useful tools for representing irrelevance or independency relationships, and they may be used to e ciently perform reasoning tasks. Singly connected networks are important speci® c cases where there is no more than one undirected path connecting each pair of variables. The aim of this paper is to investigate the kind of properties that a dependency model must verify in order to be equivalent to a singly connected graph structure, as a way of driving automated discovery and construction of singly connected networks in data. Th e main results are the characterizations of those dependency models which are isomorphic to singly connected graphs (either via the d-separation criterion for directed acyclic graphs or via the separation criterion for undirected graphs), as well as the development of e cient algorithms for learning singly connected graph representations of dependency models.Keywords : graphical models, independency relationships, learning algorithms, singly connected networks. 14 February 1997 ; revision received 11 June 1997 ; accepted 12 June 1997 1. Introduction Graphical models have become common knowled ge representation tools capable of e ciently representing and handling independency relationships as well as uncertainty in our knowledge. They comprise a qualitative and a quantitative component. The qualitative component is a graph displaying dependency} independency relationships : the absence of some links means the existence of certain conditional independency relationships between variables, and the presence of links may represent the existence of direct dependency relationships (if a causal interpretation is given, then the (directed) links signify the existence of direct causal in¯uences between the linked variables). Th is is important because an appropriate use of independency or irrelevance relationships is crucial for the management of information, since independency can modularize knowledge in such a way that we only need to consult the pieces of information relevant to the speci® c question in which we are interested, instead of having to explore a whole knowledge base. Th e quantitative component is a collection of numerical parameters, usually conditional probabilities, which give idea of the strength of the dependencies and measure our uncertainty. Therefore, Received

show abstract

Stochastic independence, causal independence, and shieldability

Cited by 143 publications

References 1 publication

Graphoids over Counterfactuals

Graphoids over Counterfactuals

Bayesian Networks and the Problem of Unreliable Instruments

Independency relationships and learning algorithms for singly connected networks

Contact Info

Product

Resources

About