A survey of cross-validation procedures for model selection
Citations
2,455 citations
1,533 citations
998 citations
Cites methods or result from "A survey of cross-validation proced..."
...…dependence of the validation with the training data will favour overfitted models (Larimore and Mehra 1985, Hawkins 2004), resulting in artificially small error estimates and thus overly optimistic estimates of model performance (Mosteller and Tukey 1977, Olden et al. 2002, Arlot and Celisse 2010)....
[...]
...When systematically compared with random data splits, they consistently demonstrate larger errors in predictions (Burman et al. 1994, Arlot and Celisse 2010, Lieske and Bender 2011, Roberts and Hamann 2012a, Wenger and Olden 2012, Bahn and McGill 2013, Radosavljevic and Anderson 2014)....
[...]
870 citations
Cites methods from "A survey of cross-validation proced..."
...One popular way to estimate the accuracy of a machine learning system when there is a limited dataset is to use the cross-validation technique (38,39)....
[...]
644 citations
References
40,785 citations
"A survey of cross-validation proced..." refers background in this paper
...Let us finally mention the Lasso (Tibshirani, 1996) and other ℓ1 penalization procedures, which have recently attracted much attention (see for instance Hesterberg et al., 2008)....
[...]
...Note that regularized least-squares algorithms such as the Lasso, ridge regression and spline smoothing also are leastsquares estimators, the model S being some ball of a (data-dependent) radius for the L1 (resp....
[...]
38,681 citations
26,531 citations
"A survey of cross-validation proced..." refers background in this paper
...In the context of statistical learning, Vapnik and Chervonenkis (1974) proposed the structural risk minimization approach (see also Vapnik, 1982, 1998)....
[...]
15,424 citations