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

The training of deep neural network classifiers results in decisionboundaries which geometry is still not well understood. This is in directrelation with classification problems such as so called adversarial examples.We introduce zoNNscan, an index that is intended to inform on the boundaryuncertainty (in terms of the presence of other classes) around one given inputdatapoint. It is based on confidence entropy, and is implemented throughsampling in the multidimensional ball surrounding that input. We detail thezoNNscan index, give an algorithm for approximating it, and finally illustrateits benefits on four applications, including two important problems for theadoption of deep networks in critical systems: adversarial examples and cornercase inputs. We highlight that zoNNscan exhibits significantly higher valuesthan for standard inputs in those two problem classes.

zoNNscan : a boundary-entropy index for zone inspection of neural models