A NEW MEASURE OF RANK CORRELATION

By M. Kendall. 1938

How researchers see this paper

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

In Statistics and Probability

Citations in the corpus

GreedyNAS: Towards Fast One-Shot NAS With Greedy Supernet (2020, methodology): “In Table 3, we report the widely-used Spearman rho [25] and Kendall tau [13] rank correlation coefficient, which are in the range [0, 1] and larger values mean stronger correlation.”
DSNAS: Direct Neural Architecture Search Without Parameter Retraining (2020, methodology): “Since the validity of the searching in two-stage NAS relies on a high correlation in the performance of searching stage and evaluation stage, we check this assumption with a ranking correlation measure, Kendall Tau metric τ [10].”
FairNAS: Rethinking Evaluation Fairness of Weight Sharing Neural Architecture Search (2019, background): “A recent work [20] evaluates the search phase based on Kendall Tau [10], which measures the ranking relation between models during the search phase and fully trained ones.”
Evaluating the Search Phase of Neural Architecture Search (2019, methodology): “As a correlation measure, we make use of the Kendall Tau (τ ) metric [11]: a number in the range [-1, 1] with the following properties: • τ = −1: Maximum disagreement.”
Bayesian Neural Word Embedding (2016, methodology): “ when applying the similarities from Eqs. (25) and (26), instead of the cosine similarity. 8 In order to compare between the BSG and SG models, we compute for each model the Spearman [17] and Kendall [18] rank correlation coefficients with respect to the ground truth. The word analogy task [8] is es”
Unsupervised Inductive Whole-Graph Embedding by Preserving Graph Proximity (2019, methodology): “For each dataset, we split it into training, validation, and testing sets by 6:2:2, and report the averaged Mean Squared Error (mse), Kendall’s Rank Correlation Coefficient (τ ) (Kendall, 1938), and Precision at 10 (p@10) to test the ranking performance.”
Unsupervised Inductive Graph-Level Representation Learning via Graph-Graph Proximity (2019, methodology): “For each dataset, we split it into training, validation, and testing sets by 6:2:2, and report the averaged Mean Squared Error (mse), Kendall’s Rank Correlation Coefficient (τ ) [Kendall, 1938], and Precision at 10 (p@10) to test the ranking performance.”
Meta-Learning: A Survey (2018, neutral): “(2015b) do this using a task similarity measure based on the Kendall tau rank correlation coefficient (Kendall, 1938) between the ranks obtained by ranking configurations θi using Pi,j and Pi,new, respectively.”
Fantastic Generalization Measures and Where to Find Them (2019, methodology): “We use Kendall’s rank coefﬁcient τ (Kendall, 1938) to capture to what degree such consistency holds among the elements of T .”
A Systematic Classification of Knowledge, Reasoning, and Context within the ARC Dataset (2018, background): “The Kemeny voting rule minimizes the Kendall Tau (Kendall, 1938) (ﬂip) distance between the output ordering and the ordering of all annotators.”
Measuring Model Biases in the Absence of Ground Truth (2021, methodology): “We use the Sørensen-Dice Coefficient (SDC), which has the commonly-used F1-score as one of its variants; the Jaccard Index (JI), a common metric in Computer Vision also known as Intersection Over Union (IOU); Log-Likelihood Ratio (LLR), a classic flexible comparison approach; Kendall Rank Correlatio”
A comparative evaluation of interactive segmentation algorithms (2010, positive): “For the comparison, we computed two measures of correlation between measured accuracy and perceived accuracy, specifically: the (Pearson product-moment) correlation coefficient, and Kendall’s tau [28] rank correlation coefficient.”
BARTScore: Evaluating Generated Text as Text Generation (2021, methodology): “3 illustrates Kendall’s Tau correlation of diverse metrics on different language pairs.”
A new rank correlation coefficient for information retrieval (2008, methodology): “Two of the most commonly used rank correlation statistics are Kendall’s τ [7] and Spearman rank correlation coefficient [15].”
Influence functions of the Spearman and Kendall correlation measures (2010, methodology): “Another nonparametric correlation measure based on signs is Kendall’s correlation (Kendall, 1938), given by rK = 2 n(n−1) ∑i j sign((xi− x j)(yi− y j)) .”
A Class of Statistics with Asymptotically Normal Distribution (1948, background): “A CLASS OF STATISTICS 319 U(') has been termed the grade of the continuous variable X(i), and in the general case PC) W(X(t) may be called the grade of X(i) (cf., for instance, G. U. Yule and M. G. Kendall [22, p. 150]).”