Online Federated Learning over decentralized networks

Online Federated Learning refers to the online optimization that is distributed over a decentralized network while still seek for training high-quality models. It allows each node to perform local operation and only contacts with its immediate neighbors, liberating it from the control of the `master...

Full description

Saved in:
Bibliographic Details
Main Author: Zhang, Chi
Other Authors: Soh Yeng Chai
Format: Theses and Dissertations
Language:English
Published: 2019
Subjects:
Online Access:https://hdl.handle.net/10356/106445
http://hdl.handle.net/10220/47928
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-106445
record_format dspace
spelling sg-ntu-dr.10356-1064452020-11-01T05:00:21Z Online Federated Learning over decentralized networks Zhang, Chi Soh Yeng Chai Interdisciplinary Graduate School (IGS) DRNTU::Engineering::Electrical and electronic engineering Online Federated Learning refers to the online optimization that is distributed over a decentralized network while still seek for training high-quality models. It allows each node to perform local operation and only contacts with its immediate neighbors, liberating it from the control of the `master' node. The computation and communication are totally decentralized, avoiding the traffic congestion and network coordination problems that are inevitable to most centralized distributed optimization. Our work addresses several critical problems and their corresponding solutions to make OFL more practical and more efficient in real-scenarios: (a) sampling technique to replace the costly deterministic communication cost with a stochastic strategy; (b) developing and analyzing an algorithm to seek for the optimal saddle point for decentralized online convex-concave problems, and therefore providing solutions for constrained decentralized optimization; (c) avoiding the challenges of “Pareto optimality” when the optimal values for models may only be similar rather than identical. Doctor of Philosophy 2019-03-28T08:26:37Z 2019-12-06T22:11:57Z 2019-03-28T08:26:37Z 2019-12-06T22:11:57Z 2018 Thesis Zhang, C. (2018).Online Federated Learning over decentralized networks. Doctoral thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/106445 http://hdl.handle.net/10220/47928 10.32657/10220/47928 en 123 p. application/pdf
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic DRNTU::Engineering::Electrical and electronic engineering
spellingShingle DRNTU::Engineering::Electrical and electronic engineering
Zhang, Chi
Online Federated Learning over decentralized networks
description Online Federated Learning refers to the online optimization that is distributed over a decentralized network while still seek for training high-quality models. It allows each node to perform local operation and only contacts with its immediate neighbors, liberating it from the control of the `master' node. The computation and communication are totally decentralized, avoiding the traffic congestion and network coordination problems that are inevitable to most centralized distributed optimization. Our work addresses several critical problems and their corresponding solutions to make OFL more practical and more efficient in real-scenarios: (a) sampling technique to replace the costly deterministic communication cost with a stochastic strategy; (b) developing and analyzing an algorithm to seek for the optimal saddle point for decentralized online convex-concave problems, and therefore providing solutions for constrained decentralized optimization; (c) avoiding the challenges of “Pareto optimality” when the optimal values for models may only be similar rather than identical.
author2 Soh Yeng Chai
author_facet Soh Yeng Chai
Zhang, Chi
format Theses and Dissertations
author Zhang, Chi
author_sort Zhang, Chi
title Online Federated Learning over decentralized networks
title_short Online Federated Learning over decentralized networks
title_full Online Federated Learning over decentralized networks
title_fullStr Online Federated Learning over decentralized networks
title_full_unstemmed Online Federated Learning over decentralized networks
title_sort online federated learning over decentralized networks
publishDate 2019
url https://hdl.handle.net/10356/106445
http://hdl.handle.net/10220/47928
_version_ 1683494325622669312