Statistical learning and selective inference

By Jonathan E. Taylor, Robert Tibshirani. 2015

400 citations recorded by the source corpus.

Subfields: Analytical Chemistry, Statistics and Probability

How researchers see this paper

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

In Analytical Chemistry

In Statistics and Probability

Citations in the corpus

Preserving Statistical Validity in Adaptive Data Analysis (2014, methodology): “es or model selection followed by testing and are optimized for speciﬁc inference procedures (the literature is too vast to adequately cover here, see Ch. 7 in [HTF09] for a textbook introduction and [TT15] for a survey of some recent work). In contrast, our framework addresses multiple stages of ad”
Asymptotically Honest Confidence Regions for High Dimensional Parameters by the Desparsified Conservative Lasso (2014, methodology): “tinguishable. Berk et al. (2013) propose a conservative post selection inference method. The idea is simultaneous inference in all models’ submodels and this results in very wide conﬁdence intervals. Taylor and Tibshirani (2015) discuss a practical way of taking into account the model selection’s eﬀ”
A Minimax Theory for Adaptive Data Analysis (2016, methodology): “, Van der Vaart (2000)) with respect to X’s distribution and X consists of iid draws from a common distribution with sample size n→∞.”
Bayesian Workflow (2020, background): “Taylor and Tibshirani (2015) warn of the problem of inference conditional on having “searched for the strongest associations.””
Three principles of data science: predictability, computability, and stability (PCS) (2018, result): “We compared the above PCS stability scores with asymptotic normality results applied to features selected by lasso and selective inference (68).”
Controlling Bias in Adaptive Data Analysis Using Information Theory (2015, background): “One recent line of work [9, 18] proposes a framework for assigning significance and confidence intervals in selective inference, where model selection and significance testing are performed on the same dataset.”
Optimal Inference After Model Selection (2014, methodology): “e results to the case of unknown ˙2, and to arbitrary exponential families after arbitrary selection events. Since the initial appearance of this work, it has been applied in many other settings; see Taylor and Tibshirani (2015) for a recent review. Other works have viewed selective inference as a m”
How Much Does Your Data Exploration Overfit? Controlling Bias via Information Usage (2015, background): “One recent line of work [8], [9] proposes a framework for assigning significance and confidence intervals in selective inference, where model selection and significance testing are performed on the same dataset.”
Why adaptively collected data have negative bias and how to correct for it (2017, methodology): “This conditioning approach is motivated by the recent successes of selective inference, which have been applied, for example, to debias the confidence intervals of the Lasso-selected features by conditioning on the Lasso algorithm [Taylor and Tibshirani, 2015].”
Predictability of Week-3–4 Average Temperature and Precipitation over the Contiguous United States (2017, methodology): “This step is tantamount to a model selection problem and is one of the most challenging problems in statistics (Fukunaga 1990; Hastie et al. 2003; Taylor and Tibshirani 2015).”
Variable selection – A review and recommendations for the practicing statistician (2018, methodology): “Recently the problem of performing valid inference on the regression coefficients after model selection by the LASSO was investigated (Taylor & Tibshirani, 2015) and a solution was proposed.”
Moving beyond regression techniques in cardiovascular risk prediction: applying machine learning to address analytic challenges (2016, background): “of specified parameters, they can be hypothesis generating and may help detect which factors are worthy of further study.(42) Missing data”
Generalization in Adaptive Data Analysis and Holdout Reuse (2015, background): “7 in [12] for a textbook introduction and [20] for a survey of some recent work).”