PREDICTION/ESTIMATION WITH SIMPLE LINEAR MODELS: IS IT REALLY THAT SIMPLE?

By Yuhong Yang. Econometric Theory, 2006

65 citations recorded by the source corpus.

Subfields: Control and Systems Engineering, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/c3ef554073ad7ae7993a824c9f7f31d770a81405

How researchers see this paper

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

In Control and Systems Engineering

In Statistics and Probability

Citations in the corpus

Model Selection Techniques: An Overview (2018, methodology): “…selection method, we can always ﬁnd a parameter value that is small enough so that the selection method tends to not include it (since it has to avoid over-selection), yet the parameter value is big enough so that dropping it has a detrimental effect in rate of convergence (see, e.g., [49], [90]).”
Variable Selection in Predictive Regressions (2013, methodology): “At least in the classical T > N , it is generally thought that the BIC is good if the true model is finite dimensional; otherwise the AIC finds the smallest possible model for prediction, cf. Yang (2007).”
Cross-validation for selecting a model selection procedure (2015, methodology): “We need to point out that Yang (2007b) deals with the setting that the true regression function is fixed when there are more and more observations.”
Catching up faster by switching sooner: a predictive approach to adaptive estimation with an application to the AIC–BIC dilemma (2012, background): “The only other procedure that we know of, for which somewhat related results have been shown, is a version of cross-validation that was proposed by Yang (2007a) to select between AIC and BIC in regression problems.”
Model selection bias and Freedman’s paradox (2010, background): “Burnham and Anderson 2004; Yang 2007; Wheeler and Bailer 2007; Wheeler 2009).”
Parametric or nonparametric? A parametricness index for model selection (2011, unknown): “For a sequence λn ≥ (logn)−1 and a constant d ≥ 0, let ICλn,d(k) = ‖Yn − Ŷk‖2 + λn log(n)rkσ 2 − nσ 2 + dn1/2 log(n)σ 2, when σ is known, and ICλn,d(k, σ̂ 2) = ‖Yn − Ŷk‖2 + λn log(n)rkσ̂ 2 − nσ̂ 2 + dn1/2 log(n)σ̂ 2, when σ is estimated by σ̂ .”
Model Selection (2018, background): “While this can be achieved in a pointwise asymptotic framework (Yang (2007), Ing (2007)), it is not surprising – given the preceding discussion – that it is impossible to achieve this goal in a uniform sense: No model selection procedure can simultaneously be consistent (like BIC) and minimax-rate…”
Model selection and psychological theory: A discussion of the differences between the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC) (2012, methodology): “…have been published that directly compare the utility of AIC and BIC in latent variable model selection, and all were done with zero/one loss functions (Nylund, Asparouhov, & Muthén, 2007; Lubke & Neale, 2006; F. Li, Cohen, Kim, & Cho, 2009; C. Yang & Yang, 2007; Henson, Reise, & Kim, 2007).”
Analysis of k-Fold Cross-Validation over Hold-Out Validation on Colossal Datasets for Quality Classification (2016, methodology): “When prior information on the true model is not available, a random use of model selection criteria may result in poor performance [13], [14].”