Round Based Extension Algorithm for Gossip Learning

Distributed training algorithms represent a promising direction in providing specially crafted machine learning models, personalized on more appropriate user data samples producing greater overall user experience.We propose an extension for the fully distributed, peer-to-peer based Gossip Learning a...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2020 IEEE 16th International Conference on Intelligent Computer Communication and Processing (ICCP) s. 251 - 257
Hlavní autoři: Bagoly, Szabolcs-Marton, Danescu, Radu Gabriel
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 03.09.2020
Témata:
Collaborative work
Federated Learning
Gossip Learning
Mobile handsets
on-device training
Peer-to-peer computing
Protocols
Servers
Task analysis
Training
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Distributed training algorithms represent a promising direction in providing specially crafted machine learning models, personalized on more appropriate user data samples producing greater overall user experience.We propose an extension for the fully distributed, peer-to-peer based Gossip Learning algorithm, enhancing it with additional memory for storing local caches of model updates. In our simulations, the extension algorithm achieved better performance for both accuracy and convergence time. One inherent advantage of the extension is that it makes the algorithm better suited for deployment on mobile devices. We also examine some technical aspects of deploying a Gossip Learning framework onto mobile devices, defining some basic building blocks, followed by technical implementation solutions for the Android operating system.
DOI:10.1109/ICCP51029.2020.9266197