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

Current transfer learning methods are mainly based on finetuning a pretrainedmodel with target-domain data. Motivated by the techniques from adversarialmachine learning (ML) that are capable of manipulating the model prediction viadata perturbations, in this paper we propose a novel approach, black-boxadversarial reprogramming (BAR), that repurposes a well-trained black-box MLmodel (e.g., a prediction API or a proprietary software) for solving differentML tasks, especially in the scenario with scarce data and constrainedresources. The rationale lies in exploiting high-performance but unknown MLmodels to gain learning capability for transfer learning. Using zeroth orderoptimization and multi-label mapping techniques, BAR can reprogram a black-boxML model solely based on its input-output responses without knowing the modelarchitecture or changing any parameter. More importantly, in the limitedmedical data setting, on autism spectrum disorder classification, diabeticretinopathy detection, and melanoma detection tasks, BAR outperformsstate-of-the-art methods and yields comparable performance to the vanillaadversarial reprogramming method requiring complete knowledge of the target MLmodel. BAR also outperforms baseline transfer learning approaches by asignificant margin, demonstrating cost-effective means and new insights fortransfer learning.

Transfer Learning without Knowing: Reprogramming Black-box Machine Learning Models with Scarce Data and Limited Resources