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

To prevent the leakage of private information while enabling automatedmachine intelligence, there is an emerging trend to integrate federatedlearning and Neural Architecture Search (NAS). Although promising as it mayseem, the coupling of difficulties from both two tenets makes the algorithmdevelopment quite challenging. In particular, how to efficiently search theoptimal neural architecture directly from massive non-iid data of clients in afederated manner remains to be a hard nut to crack. To tackle this challenge,in this paper, by leveraging the advances in proxy-less NAS, we propose aFederated Direct Neural Architecture Search (FDNAS) framework that allowshardware-aware NAS from decentralized non-iid data of clients. To further adaptfor various data distributions of clients, inspired by meta-learning, a clusterFederated Direct Neural Architecture Search (CFDNAS) framework is proposed toachieve client-aware NAS, in the sense that each client can learn a tailoreddeep learning model for its particular data distribution. Extensive experimentson real-world non-iid datasets show state-of-the-art accuracy-efficiencytrade-offs for various hardware and data distributions of clients. Our codeswill be released publicly upon paper acceptance.

FDNAS: Improving Data Privacy and Model Diversity in AutoML