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

To use machine learning in high stakes applications (e.g. medicine), we needtools for building confidence in the system and evaluating whether it isreliable. Methods to improve model reliability often require new learningalgorithms (e.g. using Bayesian inference to obtain uncertainty estimates). Analternative is to audit a model after it is trained. In this paper, we describeresampling uncertainty estimation (RUE), an algorithm to audit the pointwisereliability of predictions. Intuitively, RUE estimates the amount that aprediction would change if the model had been fit on different training data.The algorithm uses the gradient and Hessian of the model's loss function tocreate an ensemble of predictions. Experimentally, we show that RUE moreeffectively detects inaccurate predictions than existing tools for auditingreliability subsequent to training. We also show that RUE can create predictivedistributions that are competitive with state-of-the-art methods like MonteCarlo dropout, probabilistic backpropagation, and deep ensembles, but does notdepend on specific algorithms at train-time like these methods do.

Can You Trust This Prediction? Auditing Pointwise Reliability After Learning