AI Paper Atlas

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

Global Convergence of Federated Learning for Mixed Regression

Lili Su, Jiaming Xu, Pengkun Yang

NeurIPS 2022 · Federated learning and privacy › Communication-efficient FL · approach: Client participation analysis · 0 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
Use FedMD followed by iterative FedX with client clusteringClustered federated mixed regression · Mixed regression, cluster modelscluster model estimation error: d(\hat{\theta},\theta^*)\lesssim \frac{\sigma \kappa}{\rho}\nu\log\frac{e}{\nu}Improved
Use anchor-client moment descent to initialize the cluster modelsTwo-phase algorithm · Mixed regression, anchor clientsmaximum anchor model estimation error: \sup_{i \in H} \norm{\theta_{i,T} - \theta^*_{z_i}} \le \epsilon \DeltaImproved
Use iterative clustering and FedAvg or FedProx to refine modelsTwo-phase algorithm · Mixed regression, Phase 2global model estimation error: d(\theta_t,\theta^*) \le \left(1-C_1 s\eta\rho/\kappa\right)d(\theta_{t-1},\theta^*) + C_2 s\eta\sigma\nu\log\frac{e}{\nImproved
Use FedAvg with a relatively large number of local steps while keeping update stability close to 1FedX+clustering · Mixed regression, Phase 2 communicationcommunication costImproved
Allow clients with only O(1) local data points in the fine-tuning phaseTwo-phase algorithm · Mixed regression, unbalanced client dataglobal convergenceImproved

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