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

Due to user privacy and regulatory restrictions, federate learning (FL) isproposed as a distributed learning framework for training deep neural networks(DNN) on decentralized data clients. Recent advancements in FL have appliedNeural Architecture Search (NAS) to replace the predefined one-size-fit-all DNNmodel, which is not optimal for all tasks of various data distributions, withsearchable DNN architectures. However, previous methods suffer from expensivecommunication cost rasied by frequent large model parameters transmissionbetween the server and clients. Such difficulty is further amplified whencombining NAS algorithms, which commonly require prohibitive computation andenormous model storage. Towards this end, we propose FAQS, an efficientpersonalized FL-NAS-Quantization framework to reduce the communication costwith three features: weight-sharing super kernels, bit-sharing quantization andmasked transmission. FAQS has an affordable search time and demands verylimited size of transmitted messages at each round. By setting differentpersonlized pareto function loss on local clients, FAQS can yield heterogeneoushardware-aware models for various user preferences. Experimental results showthat FAQS achieves average reduction of 1.58x in communication bandwith perround compared with normal FL framework and 4.51x compared with FL+NASframwork.

FAQS: Communication-efficient Federate DNN Architecture and Quantization Co-Search for personalized Hardware-aware Preferences