AI Paper Atlas

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

DA-DFGAS:Differentiable Federated Graph Neural Architecture Search with Distribution-Aware Attentive Aggregation

Zhaowei Liu, Yihao Jiang, Rufei Gao, Jinglei Liu, Dong Yang

AAAI 2026 · Federated learning and privacy › Personalized and heterogeneous FL · approach: Federated semi-supervised and graph learning

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 distribution-aware aggregation (vs data-volume weighting) (ablation)DA-DFGAS · DBLPaccuracy (%): 58.67 -> 59.36Improved
add distribution-aware aggregation (vs data-volume weighting) (ablation)DA-DFGAS · SBM PATTERNaccuracy (%): 86.77 -> 87.35Improved
add system-heterogeneity handling (FedNova normalization and selection) (ablation)DA-DFGAS · SBM PATTERNcommunication rounds: 144 -> 118Improved
add proximal regularization to local objective (ablation)DA-DFGAS · DBLPcommunication rounds: 172 -> 97Improved
add proximal regularization to local objective (ablation)DA-DFGAS · SBM PATTERNaccuracy (%): 86.31 -> 87.35Improved
differentiable search instead of RL-based searchfederated GNN NAS · SBM PATTERNsearch time (h): 1.8 -> 1.2Improved
add attention mechanism to aggregationDA-DFGAS · Table 4 datasetsparameters: about 5-10% more parametersHurt
set similarity scaling factor tau to 1.0 (vs too small or too large) (ablation)DA-DFGAS · CoraaccuracyMixed
set architecture smoothing constant T around 60 (ablation)DA-DFGAS · CoraaccuracyMixed
use DA-DFGAS aggregation instead of FedProx/FedNova on GCN backboneFedGNN with fixed GCN backbone · SBM CLUSTERaccuracy (%): 59.67 -> 60.56Improved
DA-DFGAS vs FED-PUB on DBLP in Table 3 vs Table 5DA-DFGAS · DBLPaccuracy (%): 59.36 (Table 3) vs 58.61 (Table 5, GCN backbone)Mixed

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