concept

Kendall’s Tau

Also known as: Kendall’s tau correlation

Facts (11)

Sources

The construction and refined extraction techniques of knowledge ... nature.com Nature Feb 10, 2026 9 facts

formulaThe formula for Kendall’s Tau utilizes four variables: C (number of concordant pairs where relative order is the same in both rankings), D (number of discordant pairs), T (number of tied pairs in the first ranking), and U (number of tied pairs in the second ranking).

claimIn threat assessment tasks, Kendall’s Tau evaluates the consistency between model-generated threat rankings and reference rankings by comparing the relative order of items.

measurementIn threat assessment, there is a 0.01 divergence in Kendall’s Tau between non-desensitized and desensitized data.

claimThe evaluation metrics used in the study include BERTScore for automated scoring and Kendall’s Tau for ranking tasks.

claimKendall’s Tau is a statistical measure used to assess the ordinal association between two ranked variables, with values ranging from -1 (perfect negative correlation) to 1 (perfect positive correlation), where 0 indicates no correlation.

measurementExcluding the Task-Adaptive LoRA (TA-LoRA) module from the knowledge graph construction framework resulted in a decrease of 0.08 in Kendall’s Tau for threat assessment tasks.

procedureThe experimental evaluation of the DeepSeek-R1 70B LoRA model uses BERTScore for knowledge question answering, an overall score for tactical planning, Kendall’s Tau for threat assessment, and privacy scores of k-anonymity ≥ 5 and l-diversity ≥ 2.

procedureThe evaluation framework for multi-task performance comparison utilizes BERTScore for automated scoring, human evaluation, and the Kendall’s Tau ranking correlation coefficient for assessing threat assessment tasks.

measurementIn threat assessment tasks, the LoRA fine-tuned model achieved a Kendall’s Tau value of 0.92.

Adversarial testing of global neuronal workspace and ... - Nature nature.com Nature Apr 30, 2025 2 facts

procedureStatistical testing for MEG data was performed at the participant level, testing whether the correlation between data and theory-predicted models was greater than zero using Kendall’s tau, followed by a Mann–Whitney U rank test to compare theories.

procedureTo test GNWT and IIT predictions, researchers correlated subsampled matrices to model matrices predicted by the theories using Kendall’s tau correlation.