The International Journal of Biostatistics Statistical Inference for Variable Importance

By M. J. van der Laan. 2011

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

All Models are Wrong, but Many are Useful: Learning a Variable’s Importance by Studying an Entire Class of Prediction Models Simultaneously (2018, methodology): “The label of “variable importance” measure has been broadly used to describe approaches for either inference (van der Laan, 2006; Dı́az et al., 2015; Williamson et al., 2017) or prediction.”
Doubly robust uniform confidence band for the conditional average treatment effect function (2016, methodology): “Moreover, a uniform confidence band is not examined in van der Laan (2006).”
Towards optimal doubly robust estimation of heterogeneous causal effects (2020, background): “…robust CATE estimator in the next section, the result should also be of independent interest in other problems involving regression with estimated or imputed outcomes [Ai and Chen, 2003, Bibaut and van der Laan, 2017, Fan and Gijbels, 1994, Kennedy et al., 2017, Rubin and van der Laan, 2005, 2006].”
Correlation and variable importance in random forests (2013, background): “The main contribution on this topic is due to Zhu et al. (2012). In this last paper, the algorithm of reinforcement learning trees is introduced and analysed.”
Nonparametric variable importance assessment using machine learning techniques (2020, unknown): “…regression (Doksum and Samarov, 1995; Yao et al. , 2005; Huang and Chen, 2008); the marginalized mean difference, 𝐸 𝑃 0 {𝐸 𝑃 0 (𝑌 ∣ 𝑋 = 𝑥 1 , 𝑊) − 𝐸 𝑃 0 (𝑌 ∣ 𝑋 = 𝑥 0 , 𝑊)} (van der Laan, 2006; Chambaz et al. , 2012; Sapp et al. , 2014), where 𝑥 1 and 𝑥 0 are two meaningful…”
Causal Discovery with Attention-Based Convolutional Neural Networks (2019, background): “(If the last assumption is violated, the variable importance measures can still be applied, and subsequent experiments can determine until what degree the variable importance is causal [54].”
Semiparametric Counterfactual Density Estimation (2021, methodology): “In standard regression and density estimation problems, simple cross-validation procedures are available for this task; however, because our goal is estimation of a more nuanced counterfactual density, these require some refinement, in the same spirit as van der Laan and Dudoit [2003]. Thus in this ”
Zeileis Conditional Variable Importance for Random Forests (2015, background): “, [6] for assumptions necessary for interpreting the importance of a variable as a causal effect).”
All Models are Wrong but many are Useful: Variable Importance for Black-Box, Proprietary, or Misspecified Prediction Models, using Model Class Reliance (2018, methodology): “The label of “variable importance” measure has been broadly used to describe approaches for either inference (van der Laan, 2006; Díaz et al., 2015; Williamson et al., 2017) or prediction.”