AI Paper Atlas

Research map / Federated learning and privacy / Personalized and heterogeneous FL

LESS-VFL: Communication-Efficient Feature Selection for Vertical Federated Learning

Timothy Castiglia, Yi Zhou, Shiqiang Wang, Swanand Kadhe, Nathalie Baracaldo, Stacy Patterson

ICML 2023 · Federated learning and privacy › Personalized and heterogeneous FL · approach: Vertical federated learning · 3 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
add embedding component selection stage (LESS-VFL vs. local lasso) (ablation)LESS-VFL · Table 2 communication cost, MIMIC/Activity/Phishing/Ginacommunication cost to reach target accuracy and spurious removal: e.g., Phishing 8.10 -> 3.99 MB; Activity 26.56 -> 21.17 MBImproved
LESS-VFL instead of group lasso applied directly to VFLfeature selection in VFL · Table 2 communication cost across five datasetscommunication cost (MB): Phishing 95.22 -> 3.99 MB (~20x lower)Improved
LESS-VFL with heterogeneous (uneven) Gaussian noise distribution across partiesLESS-VFL · Table 3 Sylva and Ginacommunication cost (MB): Sylva 5.62 both; local lasso ties LESS-VFLNo clear effect
add embedding component selection (LESS-VFL vs local lasso) under uneven feature partitions (ablation)LESS-VFL · Table 3 MIMIC-III and Sylvacommunication cost (MB): MIMIC-III 7.17 both; Sylva 5.62 bothNo clear effect
LESS-VFL vs. group lasso under uneven feature partitionsLESS-VFL · Table 3 Activitycommunication cost (MB): 187.39 -> 15.76 MBImproved

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