Testing Statistical Hypotheses

By George Casella, Ingram Olkin, Stephen E. Fienberg. 2005

2,996 citations recorded by the source corpus.

Subfields: Artificial Intelligence, 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 Artificial Intelligence

In Statistics and Probability

Citations in the corpus

On Buggy Resizing Libraries and Surprising Subtleties in FID Calculation (2021, methodology): “Commonly-used ones include log-likelihood [28, 41], density estimate with Parzen window [28], Inception Score [64], Perceptual Path Length [37], Fréchet Inception Distance (FID) [30], Classification Accuracy Score and its early variants [60, 64], Classifier Twosample Tests [44, 47], precision and re”
Practical Black-Box Attacks against Machine Learning (2016, methodology): “started explaining why adversarial samples generalize and transfer between different architectures [15, 38]. Here, we build an intuition behind transferability based on statistical hypothesis testing [26] and an analysis of DNN cost gradient sign matrices. A formal treatment of transferability is le”
Recent Progress on Generative Adversarial Networks (GANs): A Survey (2019, methodology): “In addition, the Classiﬁer Two-sample Tests (C2ST) [40] is another metric based on training a binary classiﬁer.”
Fair Regression: Quantitative Definitions and Reduction-based Algorithms (2019, methodology): “The difference between CDFs is measured in the (cid:96) ∞ norm corresponding to the Kolmogorov-Smirnov statistic (Lehmann & Romano, 2006).”
Kernel‐based tests for joint independence (2016, methodology): “1) to prove the assumptions of Lehmann and Romano (2005), theorem 15.2.3. Unfortunately, we could not extend the results in Romano (1989) from Vapnik–Chervonenkis classes of sets to the required classes of functions.”
A Kernelized Stein Discrepancy for Goodness-of-fit Tests (2016, methodology): “Re-views of classical goodness-of-ﬁt tests can be found in e.g., Lehmann & Romano (2006), where most methods have computational difﬁculty for unnormalized distributions.”
Testing for Homogeneity with Kernel Fisher Discriminant Analysis (2007, methodology): “Popular procedures are the two-sample Kolmogorov-Smirnov tests or the Cramer-Von Mises tests (Lehmann and Romano, 2005), that have been frequently used to address these issues, at least for low-dimensional data.”
A tutorial on conformal prediction (2007, background): “, zn−1) can be framed as an application of the widely accepted Neyman-Pearson theory for hypothesis testing and confidence intervals [17].”
Causal Inference on Discrete Data Using Additive Noise Models (2009, methodology): “In our implementation we only use Person’s χ 2 test (e.g. E. L. Lehmann (2005)), which is most commonly used.”
Algorithmic Fairness (2020, methodology): “One is an adjusted demographic parity measure, which requires the predictor to be independent of the sensitive attribute as measured by the cumulative distribution function (CDF) of the protected group compared to the CDF of the general population [5] using the Kolmogorov-Smirnov statistic [90].”
Sampling From A Manifold (2012, background): “…2+θ3y 3+θ4y 4 , 0 ≤ y ≤ 1 This test (and its modifications by David (1939); Barton (1953, 1956)) has been shown to have a good empirical track record and comes in for repeated favorable mention in Lehman and Romano’s survey of testing goodness of fit (Lehmann and Romano, 2005, chapter 9).”
Differentially Private Identity and Closeness Testing of Discrete Distributions (2017, background): “…sample complexity of our private closeness tester is Related Work During the past two decades, distribution property testing [3] – whose roots lie in statistical hypothesis testing [23, 22] – has received considerable attention by the computer science community, see [25, 7] for two recent surveys.”
Approximate permutation tests and induced order statistics in the regression discontinuity design (2015, background): “The proof of this result follows from standard arguments (see Lehmann and Romano, 2005, Theorem 15.2.1).”