Searcharxiv⌕ Search

arXiv subjects

A. Philip Dawid

Publications and source records attributed to A. Philip Dawid.

29 records · Page 2Linked to original sources
arXiv2014

Statistical Causality from a Decision-Theoretic Perspective

We present an overview of the decision-theoretic framework of statistical causality, which is well-suited for formulating and solving problems of determining the effects of applied causes. The approach is described in detail, and is related to and contrasted with other current formulations, such as structural equation models and potential responses. Topics and applications covered include confounding, the effect of treatment on the treated, instrumental variables, and dynamic treatment strategies.

math.ST↗
arXiv2013

Retrospective-prospective symmetry in the likelihood and Bayesian analysis of case-control studies

Prentice & Pyke (1979) established that the maximum likelihood estimate of an odds-ratio in a case-control study is the same as would be found by running a logistic regression: in other words, for this specific target the incorrect prospective model is inferentially equivalent to the correct retrospective model. Similar results have been obtained for other models, and conditions have also been identified under which the corresponding Bayesian property holds, namely that the posterior distribution of the odds-ratio be the same, whether computed using the prospective or the retrospective likelihood. Here we demonstrate how these results follow directly from certain parameter independence properties of the models and priors, and identify prior laws that support such reverse analysis, for both standard and stratified designs.

stat.ME↗
arXiv2013

A Formal Treatment of Sequential Ignorability

Taking a rigorous formal approach, we consider sequential decision problems involving observable variables, unobservable variables, and action variables. We can typically assume the property of extended stability, which allows identification (by means of G-computation) of the consequence of a specified treatment strategy if the unobserved variables are, in fact, observed - but not generally otherwise. However, under certain additional special conditions we can infer simple stability (or sequential ignorability), which supports G-computation based on the observed variables alone. One such additional condition is sequential randomization, where the unobserved variables essentially behave as random noise in their effects on the actions. Another is sequential irrelevance, where the unobserved variables do not influence future observed variables. In the latter case, to deduce sequential ignorability in full generality requires additional positivity conditions. We show here that these positivity conditions are not required when all variables are discrete.

math.ST↗
arXiv2012

Proper local scoring rules

We investigate proper scoring rules for continuous distributions on the real line. It is known that the log score is the only such rule that depends on the quoted density only through its value at the outcome that materializes. Here we allow further dependence on a finite number $m$ of derivatives of the density at the outcome, and describe a large class of such $m$-local proper scoring rules: these exist for all even $m$ but no odd $m$. We further show that for $m\geq2$ all such $m$-local rules can be computed without knowledge of the normalizing constant of the distribution.

math.ST↗
arXiv2012

Proper local scoring rules on discrete sample spaces

A scoring rule is a loss function measuring the quality of a quoted probability distribution $Q$ for a random variable $X$, in the light of the realized outcome $x$ of $X$; it is proper if the expected score, under any distribution $P$ for $X$, is minimized by quoting $Q=P$. Using the fact that any differentiable proper scoring rule on a finite sample space ${\mathcal{X}}$ is the gradient of a concave homogeneous function, we consider when such a rule can be local in the sense of depending only on the probabilities quoted for points in a nominated neighborhood of $x$. Under mild conditions, we characterize such a proper local scoring rule in terms of a collection of homogeneous functions on the cliques of an undirected graph on the space ${\mathcal{X}}$. A useful property of such rules is that the quoted distribution $Q$ need only be known up to a scale factor. Examples of the use of such scoring rules include Besag's pseudo-likelihood and Hyvärinen's method of ratio matching.

math.ST↗
arXiv2011

Probability-free pricing of adjusted American lookbacks

Consider an American option that pays G(X^*_t) when exercised at time t, where G is a positive increasing function, X^*_t := \sup_{s\le t}X_s, and X_s is the price of the underlying security at time s. Assuming zero interest rates, we show that the seller of this option can hedge his position by trading in the underlying security if he begins with initial capital X_0\int_{X_0}^{\infty}G(x)x^{-2}dx (and this is the smallest initial capital that allows him to hedge his position). This leads to strategies for trading that are always competitive both with a given strategy's current performance and, to a somewhat lesser degree, with its best performance so far. It also leads to methods of statistical testing that avoid sacrificing too much of the maximum statistical significance that they achieve in the course of accumulating data.

q-fin.PR↗
arXiv2010

Deep determinism and the assessment of mechanistic interaction between categorical and continuous variables

Our aim is to detect mechanistic interaction between the effects of two causal factors on a binary response, as an aid to identifying situations where the effects are mediated by a common mechanism. We propose a formalization of mechanistic interaction which acknowledges asymmetries of the kind "factor A interferes with factor B, but not viceversa". A class of tests for mechanistic interaction is proposed, which works on discrete or continuous causal variables, in any combination. Conditions under which these tests can be applied under a generic regime of data collection, be it interventional or observational, are discussed in terms of conditional independence assumptions within the framework of Augmented Directed Graphs. The scientific relevance of the method and the practicality of the graphical framework are illustrated with the aid of two studies in coronary artery disease. Our analysis relies on the "deep determinism" assumption that there exists some relevant set V - possibly unobserved - of "context variables", such that the response Y is a deterministic function of the values of V and of the causal factors of interest. Caveats regarding this assumption in real studies are discussed.

stat.ME↗
arXiv2010

Insuring against loss of evidence in game-theoretic probability

We consider the game-theoretic scenario of testing the performance of Forecaster by Sceptic who gambles against the forecasts. Sceptic's current capital is interpreted as the amount of evidence he has found against Forecaster. Reporting the maximum of Sceptic's capital so far exaggerates the evidence. We characterize the set of all increasing functions that remove the exaggeration. This result can be used for insuring against loss of evidence.

math.ST↗
arXiv2010

Identifying the consequences of dynamic treatment strategies: A decision-theoretic overview

We consider the problem of learning about and comparing the consequences of dynamic treatment strategies on the basis of observational data. We formulate this within a probabilistic decision-theoretic framework. Our approach is compared with related work by Robins and others: in particular, we show how Robins's 'G-computation' algorithm arises naturally from this decision-theoretic perspective. Careful attention is paid to the mathematical and substantive conditions required to justify the use of this formula. These conditions revolve around a property we term stability, which relates the probabilistic behaviours of observational and interventional regimes. We show how an assumption of 'sequential randomization' (or 'no unmeasured confounders'), or an alternative assumption of 'sequential irrelevance', can be used to infer stability. Probabilistic influence diagrams are used to simplify manipulations, and their power and limitations are discussed. We compare our approach with alternative formulations based on causal DAGs or potential response models. We aim to show that formulating the problem of assessing dynamic treatment strategies as a problem of decision analysis brings clarity, simplicity and generality.

math.ST↗
arXiv2004

Game theory, maximum entropy, minimum discrepancy and robust Bayesian decision theory

We describe and develop a close relationship between two problems that have customarily been regarded as distinct: that of maximizing entropy, and that of minimizing worst-case expected loss. Using a formulation grounded in the equilibrium theory of zero-sum games between Decision Maker and Nature, these two problems are shown to be dual to each other, the solution to each providing that to the other. Although Topsœdescribed this connection for the Shannon entropy over 20 years ago, it does not appear to be widely known even in that important special case. We here generalize this theory to apply to arbitrary decision problems and loss functions. We indicate how an appropriate generalized definition of entropy can be associated with such a problem, and we show that, subject to certain regularity conditions, the above-mentioned duality continues to apply in this extended context. This simultaneously provides a possible rationale for maximizing entropy and a tool for finding robust Bayes acts. We also describe the essential identity between the problem of maximizing entropy and that of minimizing a related discrepancy or divergence between distributions. This leads to an extension, to arbitrary discrepancies, of a well-known minimax theorem for the case of Kullback-Leibler divergence (the ``redundancy-capacity theorem'' of information theory). For the important case of families of distributions having certain mean values specified, we develop simple sufficient conditions and methods for identifying the desired solutions.

math.ST↗