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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Expanding on MacKay (1992), we argue that conventional model-based methodsfor active learning - like BALD - have a fundamental shortfall: they fail todirectly account for the test-time distribution of the input variables. Thiscan lead to pathologies in the acquisition strategy, as what is maximallyinformative for model parameters may not be maximally informative forprediction: for example, when the data in the pool set is more dispersed thanthat of the final prediction task, or when the distribution of pool and testsamples differs. To correct this, we revisit an acquisition strategy that isbased on maximizing the expected information gained about possible futurepredictions, referring to this as the Expected Predictive Information Gain(EPIG). As EPIG does not scale well for batch acquisition, we further examinean alternative strategy, a hybrid between BALD and EPIG, which we call theJoint Expected Predictive Information Gain (JEPIG). We consider using both foractive learning with Bayesian neural networks on a variety of datasets,examining the behavior under distribution shift in the pool set.

Test Distribution-Aware Active Learning: A Principled Approach Against Distribution Shift and Outliers