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

High annotation cost for training machine learning classifiers has drivenextensive research in active learning and self-supervised learning. Recentresearch has shown that in the context of supervised learning different activelearning strategies need to be applied at various stages of the trainingprocess to ensure improved performance over the random baseline. We refer tothe point where the number of available annotations changes the suitable activelearning strategy as the phase transition point. In this paper, we establishthat when combining active learning with self-supervised models to achieveimproved performance, the phase transition point occurs earlier. It becomeschallenging to determine which strategy should be used for previously unseendatasets. We argue that existing active learning algorithms are heavilyinfluenced by the phase transition because the empirical risk over the entireactive learning pool estimated by these algorithms is inaccurate and influencedby the number of labeled samples. To address this issue, we propose a novelactive learning strategy, neural tangent kernel clustering-pseudo-labels(NTKCPL). It estimates empirical risk based on pseudo-labels and the modelprediction with NTK approximation. We analyze the factors affecting thisapproximation error and design a pseudo-label clustering generation method toreduce the approximation error. We validate our method on five datasets,empirically demonstrating that it outperforms the baseline methods in mostcases and is valid over a wider range of training budgets.

NTKCPL: Active Learning on Top of Self-Supervised Model by Estimating True Coverage