Present position and potential developments: some personal views

By A. Dawid. 1984

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

A Meta-Transfer Objective for Learning to Disentangle Causal Mechanisms (2019, methodology): “Note that, in addition to its contribution to the update of the parameters, each data point xt is also used to evaluate the performance of our model so far; this is called a prequential analysis (Dawid, 1984), also corresponding to sequential cross-validation (Gingras et al.”
Speeding-Up Hoeffding-Based Regression Trees With Options (2011, methodology): “The weighted prequential error wPreqMSE( T i , t) is an improved variant of the prequential error originally proposed by Dawid (1984) and using a fading factor as proposed by Gama et al. (2009).”
Bayes factors (1995, methodology): “5 Evaluation of Methods The e cacy of a modeling strategy can be judged by how well the resulting predictive distributions predict future observations (Dawid, 1984). Madigan and Raftery (1994) measured predictive performance by splitting complete data into two subsets, one of which (typically about ”
Regression diagnostics meets forecast evaluation: conditional calibration, reliability diagrams, and coefficient of determination (2021, methodology): “While checks for probabilistic calibration have become a cornerstone of predictive distribution evaluation (Dawid, 1984; Diebold et al., 1998; Gneiting et al., 2007), both marginal and probabilistic calibration concern unconditional facets of predictive performance, which is increasingly being…”
Evaluating the Calibration of Multi-Step-Ahead Density Forecasts Using Raw Moments (2015, methodology): “If the density forecasts are calibrated correctly, the PITs are uniformly distributed over the interval (0, 1), as noted by Dawid (1984), Diebold, Gunther, and Tay (1998), and Diebold, Tay, and Wallis (1999).”
A Survey of Model Evaluation Approaches With a Tutorial on Hierarchical Bayesian Methods (2008, methodology): “The prequential method (from “sequential prediction”; Dawid, 1984, 1991, 1992; Dawid & Vovk, 1999; Skouras & Dawid, 1998) is based on the philosophy that the relevant criterion for model selection is the minimization of prediction error for unseen data.”
The Safe Bayesian - Learning the Learning Rate via the Mixability Gap (2012, background): “a prime instance of Dawid’s (1984) prequential approach to statistics, which inspired”
Online Machine Learning in Big Data Streams (2018, methodology): “Most studies of online learning determine overall loss as a sum of losses experienced by individual training examples. Based on this so-called predictive sequential, abbreviated as prequential method [64], we deﬁne online training and evaluation in the following steps: 12 Andr´as A. Bencz ur and Lev”
Comparison of non-homogeneous regression models for probabilistic wind speed forecasting (2013, unknown): “The continuous analog of the VR histogram is the probability integral transform (PIT) histogram (Dawid, 1984; Gneiting et al., 2007).”
Functional Coefficient Regression Models for Non‐linear Time Series: A Polynomial Spline Approach (2004, methodology): “The approach we used to evaluate the density forecasts is based on Dawid (1984) and Diebold et al. (1998) and was applied in, for example, Clements & Smith (2000).”
Combining forecast densities from VARs with uncertain instabilities (2010, methodology): “A popular method, following Rosenblatt (1952), Dawid (1984) and Diebold et al. (1998), evaluates using the probability integral transforms (pits) of the realization of the variable with respect to the forecast densities.”
Model selection for probabilistic clustering using cross-validated likelihood (2000, unknown): “Of course, as mentioned earlier, implementing the Bayesian approach in practice involves approximation of one form or another and, indeed, cross-validation itself can be viewed as a different type of approximation in the Bayesian context (Dawid 1984).”
Model Selection and the Principle of Minimum Description Length (2001, methodology): “The reasoning we followed in deriving PMDL is identical to the prequential approach to statistics advocated by Dawid (1984, 1991). The form (18) appeared in the literature on Gaussian regression and time series analysis as the predictive least squares criterion long before the development of MDL, an”
Dark Energy Survey Year 1 Results: redshift distributions of the weak-lensing source galaxies (2017, background): “The value c(zspec) for a distribution of objects is the Probability Integral Transform (PIT, Dawid 1984; Angus 1994).”
Assessment and Propagation of Model Uncertainty (2011, methodology): “However, as Dawid (1984, 1985), Hodges (1987), and others have noted, to be fully useful an inference or prediction must also have an uncertainty assessment attached to it, and it is also important for this \giveor-take" to be accurate, because otherwise choices are made that incorporate too little…”
Bayesian Networks for Data Mining (2004, methodology): “The log marginal likelihood has the following interesting interpretation described by Dawid (1984). From the chain rule of probability, we have”