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

This paper studies a new problem, \emph{active learning with partial labels}(ALPL). In this setting, an oracle annotates the query samples with partiallabels, relaxing the oracle from the demanding accurate labeling process. Toaddress ALPL, we first build an intuitive baseline that can be seamlesslyincorporated into existing AL frameworks. Though effective, this baseline isstill susceptible to the \emph{overfitting}, and falls short of therepresentative partial-label-based samples during the query process. Drawinginspiration from human inference in cognitive science, where accurateinferences can be explicitly derived from \emph{counter-examples} (CEs), ourobjective is to leverage this human-like learning pattern to tackle the\emph{overfitting} while enhancing the process of selecting representativesamples in ALPL. Specifically, we construct CEs by reversing the partial labelsfor each instance, and then we propose a simple but effective WorseNet todirectly learn from this complementary pattern. By leveraging the distributiongap between WorseNet and the predictor, this adversarial evaluation mannercould enhance both the performance of the predictor itself and the sampleselection process, allowing the predictor to capture more accurate patterns inthe data. Experimental results on five real-world datasets and four benchmarkdatasets show that our proposed method achieves comprehensive improvements overten representative AL frameworks, highlighting the superiority of WorseNet. Thesource code will be available at \url{https://github.com/Ferenas/APLL}.

Exploiting Counter-Examples for Active Learning with Partial labels