Asymptotics of cross-validated risk estimation in estimator selection and performance assessment

By Sandrine Dudoit, M. J. van der Laan. 2005

210 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/9d0e0b55efc2d7b145693df89d458be2843865ff

How researchers see this paper

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

In Statistics and Probability

Citations in the corpus

Limits of Estimating Heterogeneous Treatment Effects: Guidelines for Practical Algorithm Design (2018, methodology): “It can be easily shown using the results in (Dudoit & van der Laan, 2005) that when using the estimator above to tune the GP hyperparameters via cross-validation, then the learned length-scale parameters will satisfy the matching condition for minimax optimality.”
Computer assisted customer churn management: State-of-the-art and future trends (2007, methodology): “Several cross-validation methods have been proposed in the literature, two examples follow [65]: ◦ V-fold cross validation—the learning set is randomly partitioned into limited datasets of equal size.”
Targeted Data Adaptive Estimation of the Causal Dose–Response Curve (2013, unknown): “Following arguments similar to those presented by Dudoit and van der Laan [21] and van der Vaart et al.”
Bayesian Nonparametric Causal Inference: Information Rates and Learning Algorithms (2017, methodology): “The empirical cross-validated risk estimate is deﬁned as The consistency of the estimator β ( K ∗ n ) follows from the results of Dudoit and van der Laan on the asymptotic performance of model selection via cross-validation for general loss functions [60].”
Cross-validation: what does it estimate and how well does it do it? (2021, methodology): “Lastly, we evaluate the nested CV procedure on real data sets from the UCI repository (Dua and Graff, 2017).”
Cross-validation pitfalls when selecting and assessing regression and classification models (2014, methodology): “Dudoit and van der Laan [8] give the asymptotic proof of selecting the tuning parameter with minimal cross-validation error in V-fold cross-validation and, therefore, provide a theoretical basis for this approach.”
Local polynomial regression on unknown manifolds (2007, background): “There are three main approaches: plug-in ([7, 16, 18], etc); the bootstrap ([3, 11, 12], etc) and cross validation ([6, 10, 22], etc).”
The International Journal of Biostatistics (2017, unknown): “In van der Laan et al. (2005) history adjusted marginal structural models were proposed as a natural extension of marginal structural models, and it was shown that the latter also imply an individualized treatment rule of interest (a so called history adjusted statically optimal treatment regimen):…”