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

In supervised learning, automatically assessing the quality of the labelsbefore any learning takes place remains an open research question. In certainparticular cases, hypothesis testing procedures have been proposed to assesswhether a given instance-label dataset is contaminated with class-conditionallabel noise, as opposed to uniform label noise. The existing theory builds onthe asymptotic properties of the Maximum Likelihood Estimate for parametriclogistic regression. However, the parametric assumptions on top of which theseapproaches are constructed are often too strong and unrealistic in practice. Toalleviate this problem, in this paper we propose an alternative path by showinghow similar procedures can be followed when the underlying model is a productof Local Maximum Likelihood Estimation that leads to more flexiblenonparametric logistic regression models, which in turn are less susceptible tomodel misspecification. This different view allows for wider applicability ofthe tests by offering users access to a richer model class. Similarly toexisting works, we assume we have access to anchor points which are provided bythe users. We introduce the necessary ingredients for the adaptation of thehypothesis tests to the case of nonparametric logistic regression andempirically compare against the parametric approach presenting both syntheticand real-world case studies and discussing the advantages and limitations ofthe proposed approach.

Hypothesis Testing for Class-Conditional Noise Using Local Maximum Likelihood