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 learning with noisy labels, the sample selection approach is very popular,which regards small-loss data as correctly labeled during training. However,losses are generated on-the-fly based on the model being trained with noisylabels, and thus large-loss data are likely but not certainly to be incorrect.There are actually two possibilities of a large-loss data point: (a) it ismislabeled, and then its loss decreases slower than other data, since deepneural networks "learn patterns first"; (b) it belongs to an underrepresentedgroup of data and has not been selected yet. In this paper, we incorporate theuncertainty of losses by adopting interval estimation instead of pointestimation of losses, where lower bounds of the confidence intervals of lossesderived from distribution-free concentration inequalities, but not lossesthemselves, are used for sample selection. In this way, we also give large-lossbut less selected data a try; then, we can better distinguish between the cases(a) and (b) by seeing if the losses effectively decrease with the uncertaintyafter the try. As a result, we can better explore underrepresented data thatare correctly labeled but seem to be mislabeled at first glance. Experimentsdemonstrate that the proposed method is superior to baselines and robust to abroad range of label noise types.

Sample Selection with Uncertainty of Losses for Learning with Noisy Labels