A Predictive Approach to Model Selection

By S. Geisser, W. Eddy. 1979

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 Bayesian model for repeated measures zero-inflated count data with application to outpatient psychiatric service use (2010, background): “…sample consists of a single observation, with y[i] denoting the response vector with the ith observation deleted, then the predictive distribution f (yi |y[i]) = ∫ f (yi |θ , y[i])π(θ |y[i]) dθ (3.1) is called the conditional predictive ordinate (CPO) for observation i (Geisser and Eddy, 1979).”
A rigorous framework for optimization of expensive functions by surrogates (1998, background): “[30], it has been defended by others as a crude form of cross-validation [19, 10].”
Bayesian Approach for Nonlinear Random Effects Models (1997, methodology): “which serves as a useful criterion for model selection (called the pseudo-Bayes factor for M1 with respect to Al12), was first used by Geisser and Eddy (1979). Finally, we present a sampling/importance resampling (SIR) method to generate Yr from its predictive distribution f(Yr I Y(r)) given in (2.”
Content-based recommendations with Poisson factorization (2014, methodology): “We use the predictive approach to evaluating model ﬁtness [7], comparing the predictive accuracy of the CTPF coordinate ascent algorithm in Figure 2 to collaborative topic regression (CTR) [21].”
Assessment and Propagation of Model Uncertainty (2011, background): “…implicitly or explicitly, in the writings of workers in at least three elds: statistics (e.g., Box and Tiao, 1962; de Finetti, 1972; Davis, 1979; Geisser and Eddy, 1979; Smith and Spiegelhalter, 1981; Stewart and Davis, 1986; Brown and Lindley, 1986; Draper et al., 1987; Hodges, 1987; Lavine,…”
Robustifying Generalized Linear Mixed Models Using a New Class of Mixtures of Multivariate Polya Trees (2009, methodology): “Indeed in the present work we have adopted the so-called Pseudo Marginal Likelihood (PML) and Pseudo-Bayes factor (PBF) developed by Geisser and Eddy (1979) and further considered by Gelfand and Dey (1994).”
Bayesian Modeling Using WinBUGS (2009, background): “The pseudo-marginal likelihood of Geisser and Eddy (1979) is defined by (1).”
Benchmark Priors for Bayesian Model Averaging (2001, background): “See also Dawid (1984) and Geisser and Eddy (1979).”
Comparison of Bayesian predictive methods for model selection (2015, methodology): “In other words, the variable searching was performed using the projection, the inner layer of cross-validation (10-fold) was used to decide the model size and the outer layer to measure the performance of the ﬁnally selected models (10-fold for Ionosphere and Sonar, LOO-CV for Ovarian and Colon, and”
Bayesian Variable Selection in Linear Regression (1988, background): “Various approaches to this kind of difficulty were suggested by Halpern (1973), Atkinson (1978), Smith and Spiegelhalter (1980), Spiegelhalter and Smith (1982), and Pericchi (1984).”
Bayesian Estimation and Model Choice in Item Response Models (2002, methodology): “The pseudo-Bayes factor (PsBF) (Geisser and Eddy, 1979) for comparing two models M1 and M2 is defined as,”
BAYES FACTORS AND MARGINAL DISTRIBUTIONS IN INVARIANT SITUATIONS (2016, background): “Precursors or variants on such use of training samples can be found in Lempers (1971), Atkinson (1978), Geisser and Eddy (1979) , Smith and Spiegelhalter (1982), and Gelfand, Dey and Chang (1992).”