Backend
Fuzzy-FedAvg: Server-Aware Client Selection for Federated Speech Emotion Recognition in Heterogeneous Environments
arXiv cs.SD0 views
arXiv:2611.08818v1 Announce Type: new Abstract: Federated speech emotion recognition (SER) is challenged by speaker-dependent, class-imbalanced data and by heterogeneous client devices. This paper proposes \textit{Fuzzy-FedAvg}, a server-aware framework in which the server learns Client Selection Weights (CSWs) through fuzzy Q-learning from global-model feedback. The same CSWs weight client accuracy, training time, and \divkpi{} in both client selection and aggregation, while a dedicated PCAS node manages stragglers through staleness-discounted deferred updates. On IEMOCAP and CREMA-D, \textit{Fuzzy-FedAvg} achieves higher unweighted average recall (UAR) than FedAvg in all configurations under severe non-IID conditions, and in 22 of 24 participation-level and 10 of 12 training-horizon configurations overall. It also outperforms FedProx and Oort, while SCAFFOLD and Power-of-Choice remain competitive, and it reduces cumulative training time by 81.2\% in a FedScale-based evaluation. The benefit is greatest under strong statistical heterogeneity and narrows as data become more balanced.
Source：https://arxiv.org/abs/2611.08818PreviousAircraftWitness: How Often Aircraft Can Locate RecordingsNextConcordance-Optimal Continuous Music Emotion Estimation: Variance Decomposition, Dual-Branch Temporal Fusion, and Conformal Prediction Intervals
Related posts
# Meshy T2: Native 3D Mesh Generation with Flow Matching
0dev.to Machine Learning
AWS Launches Bedrock Knowledge Bases Multi-Modal RAG: Native Embedding Ingestion for Graphs, Tables, and Charts
0dev.to Machine Learning
Export Your VAT Check History as a CSV for an Audit
0dev.to Tutorial
How to implement async Python that actually works in production
0dev.to Tutorial
The TCM Graduate at Majuqiao: A Day-Labor Story Followed by Feng Ge
0dev.to Career
Your Mentee Wants to Move Into Product Management
0dev.to Career