A survey of sampling from contaminated distributions

By J. Tukey. 1960

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

Sever: A Robust Meta-Algorithm for Stochastic Optimization (2018, background): “This leads us to the following conceptual question underlying the present work: Can we design training algorithms that are robust to the presence of an ε-fraction of arbitrary (and potentially adversarial) outliers? Estimation in the presence of outliers is a prototypical goal in robust statistics a”
Learning Discrete Distributions from Untrusted Batches (2017, background): “The problem of estimation and learning in the presence of contaminated or outlying data points has a history of study in the Statistics community, dating back to early work of Tukey [Tuk60] (see also the surveys [HR, HRRS11]).”
Learning from untrusted data (2016, methodology): “In this case, we can identify outliers i Ig by considering the value of fi (w̃i ) − fi (ŵi ), and Lemma 4.5 verifies that we can use this to perform outlier removal.”
Being Robust (in High Dimensions) Can Be Practical (2017, background): “Robust statistics was founded in the seminal works of (Tukey, 1960) and (Huber, 1964).”
A Data Prism: Semi-verified learning in the small-alpha regime (2018, background): “The general challenge of developing algorithms and estimators that are robust to corruptions in the input data dates back to the early work of Tukey (Tukey, 1960), and has led to a significant body of work on “Robust Statistics”, which explores a number of different models for the data corruptions, ”
Adversarial Robustness with Semi-Infinite Constrained Learning (2021, background): “While robustness has been studied in statistics for decades [23–25], this issue has been exacerbated by the opacity, scale, and non-convexity of modern learning models, such as convolutional neural network (CNNs).”
A Meta-Analysis of Gifted and Talented Identification Practices (2018, background): “Tukey (1960, 1977) espoused elimination of outliers that exist three times the interquartile range of either the 75th or 25th percentiles.”
Robust Estimation of the Correlation Coefficient: An Attempt of Survey (2011, methodology): “To illustrate the necessity in robust counterparts of the sample correlation coefficient, consider Tukey’s gross error model [18] described by the mixture of normal densities (0 ≤ ε < 0.”
Empirical Asset Pricing via Machine Learning (2018, methodology): “8Classical analyses include Box (1953), Tukey (1960), and Huber (1964).”
Image processing and analysis methods for the Adolescent Brain Cognitive Development Study (2018, methodology): “Weighting factors are calculated using Tukey’s bisquare function such that lower weights are assigned to voxels with MD values farther from the median value, relative to the dispersion values multiplied by 4.7 (Tukey, 1960).”