Large-scale inference

By Bradley Efron. 2010

819 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/44d86f3ba9167f4931dae729d814c13fbc7a0efb

How researchers see this paper

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

In Artificial Intelligence

Citations in the corpus

Machine Learning and Causal Inference : A Modular Approach to Assessing the Effects of the London Bombings of 2005 (2015, background): “11See Efron (2010) or Hastie et al. (2005), for example, on why penalized models seem to work so well compared to unpenalized models. this case, λ and α together determine a model — if λ is large, then the model will have many zero coefficients (i.e. the model will exclude many covariates or…”
Reweighted Data for Robust Probabilistic Models (2016, background): “1 Localization also relates to James-Stein shrinkage; Efron (2010) connects these dots.”
Estimation of a two‐component mixture model with applications to multiple testing (2012, methodology): “…problems (microarray analysis, neuroimaging) the p-values, obtained from the numerous (independent) hypotheses tests, are uniformly distributed on [0,1], under hypothesis H0, whereas their distribution associated with H1 is unknown; see, for example, Efron (2010) and Robin et al. (2007).”
Panning for Gold: Model-free Knockoffs for High-dimensional Controlled Variable Selection (2016, background): “Therefore, FDR control is only useful if the realized FDP is relatively concentrated around its expectation, and it is well-established (Efron, 2010, Chapter 4) that under correlations BHq can produce highly skewed FDP distributions.”
Censorship and deletion practices in Chinese social media (2012, result): “It is analogous to (1−precision) and quite different than the Type I or II error rates in classical hypothesis testing—and arguably more meaningful for large-scale inference (Efron, 2010; Storey and Tibshirani, 2003).”
Bayesian and Empirical Bayesian Forests (2015, background): “Empirical Bayes (EB) is an established framework with a successful track record in fast approximate Bayesian inference; see, e.g., Efron (2010) for an overview.”
Expectation-Maximization Gaussian-Mixture Approximate Message Passing (2012, methodology): “Since, in our approach, the prior pdf parameters are treated as deterministic unknowns, our proposed EM-GM-AMP algorithm can be classified as an “empirical-Bayesian” approach [16].”
E-values: Calibration, combination and applications (2019, neutral): “2, where in practice the true fdr( z i ) is replaced by its empirical estimate [Efron, 2010, Section 5.1].”
Variable selection using pseudo-variables (2018, methodology): “Our second illustrative example used data from leukemia study (see Efron, 2012).”
Controlling the false discovery rate via knockoffs (2014, methodology): “imated by π 0p·P{|N(0,1)| ≥t}, where π 0p is the total number of null hypotheses. For π 0 < 1, then the fraction in the deﬁnition of (3.2) is an overestimate of the FDP by a factor of (π 0)−1; see [11] and references therein. Turning to the problem of variable selection in regression, the BHq pro”
Covariate powered cross‐weighted multiple testing (2017, methodology): “imator and is related to the nonparametric properties of the empirical distribution function. In fact, it is a surprising result that FDR is controlled, despite the greedy choice of the threshold ˆt∗ [17]. 2.2. The weighted Benjamini-Hochberg procedure Despite the favourable properties of the BH pro”