AI Paper Atlas

Research map / Federated learning and privacy / Communication-efficient FL

AdaBest: Minimizing Client Drift in Federated Learning via Adaptive Bias Estimation

Farshid Varno, Marzie Saghayi, Laya Rafiee Sevyeri, Sharut Gupta, Stan Matwin, Mohammad Havaei

ECCV 2022 · Federated learning and privacy › Communication-efficient FL · approach: Client participation analysis · 36 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace FedDyn server update with AdaBest adaptive bias estimation (beta=0.96)FedDyn · CIFAR100, alpha=0.3, 10% client participationtop-1 test accuracy (%): 48.92 -> 54.51Improved
replace FedDyn with AdaBestFedDyn · CIFAR100, IID, 10% client participationtop-1 test accuracy (%): 49.60 -> 55.58Improved
replace FedDyn with AdaBestFedDyn · CIFAR100, alpha=0.03, 10% client participationtop-1 test accuracy (%): 44.24 -> 48.56Improved
replace FedDyn with AdaBestFedDyn · CIFAR100, alpha=0.3, 1% client participationtop-1 test accuracy (%): 50.53 -> 53.62Improved
replace FedDyn with AdaBestFedDyn · CIFAR100, alpha=0.03, 1% client participationtop-1 test accuracy (%): 46.27 -> 47.91Improved
replace FedDyn with AdaBestFedDyn · EMNIST-L, alpha=0.03, 1% client participationtop-1 test accuracy (%): 92.42 -> 94.49Improved
replace SCAFFOLD with AdaBestSCAFFOLD · CIFAR100, alpha=0.3, 10% client participationtop-1 test accuracy (%): 46.34 -> 54.51Improved
replace FedAvg with AdaBestFedAvg · CIFAR100, alpha=0.3, 10% client participationtop-1 test accuracy (%): 38.78 -> 54.51Improved
AdaBest versus FedDyn at full participationFedDyn · CIFAR100, alpha=0.03, 100% client participationtop-1 test accuracy (%): 46.00 -> 46.16No clear effect
AdaBest versus SCAFFOLD at full participationSCAFFOLD · CIFAR100, alpha=0.03, 100% client participationtop-1 test accuracy (%): 46.51 -> 46.16Hurt
replace FedDyn with AdaBest at full participationFedDyn · CIFAR100, IID, 100% client participationtop-1 test accuracy (%): 50.61 -> 51.33Improved
replace SCAFFOLD with AdaBest at full participationSCAFFOLD · EMNIST-L, alpha=0.3, 100% client participationtop-1 test accuracy (%): 94.53 -> 94.40Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy