English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Estimating the strength of dependency between two variables is fundamentalfor exploratory analysis and many other applications in data mining. Forexample: non-linear dependencies between two continuous variables can beexplored with the Maximal Information Coefficient (MIC); and categoricalvariables that are dependent to the target class are selected using Gini gainin random forests. Nonetheless, because dependency measures are estimated onfinite samples, the interpretability of their quantification and the accuracywhen ranking dependencies become challenging. Dependency estimates are notequal to 0 when variables are independent, cannot be compared if computed ondifferent sample size, and they are inflated by chance on variables with morecategories. In this paper, we propose a framework to adjust dependency measureestimates on finite samples. Our adjustments, which are simple and applicableto any dependency measure, are helpful in improving interpretability whenquantifying dependency and in improving accuracy on the task of rankingdependencies. In particular, we demonstrate that our approach enhances theinterpretability of MIC when used as a proxy for the amount of noise betweenvariables, and to gain accuracy when ranking variables during the splittingprocedure in random forests.

A Framework to Adjust Dependency Measure Estimates for Chance