New Algorithm Improves Federated Learning When Devices Dynamically Join or Leave Networks
Researchers have proposed HASA (Heterogeneity-Aware Subnet Allocation), a federated learning technique that assigns neural network subnet widths to edge clients based on the statistical heterogeneity of their local data rather than device constraints alone. Federated learning allows devices to collaboratively train models without sharing raw data, but existing approaches struggle to account for differences in both hardware and data distributions across clients. HASA addresses this gap and demonstrates improved accuracy for the weakest-performing clients, which matters for ensuring equitable model performance across diverse real-world deployments.
HASA is a train-only allocation rule designed for model-heterogeneous federated learning scenarios where edge devices have varying computational resources and locally distributed data that may differ significantly from one another. Rather than assigning subnet widths purely based on device capacity, HASA computes a heterogeneity score from each client's local training data and uses it to guide subnet allocation while enforcing a fixed size-weighted compute budget, enabling fair comparisons with alternative policies. In a benchmark involving next-word prediction on article titles across seven clients, HASA raised unweighted mean client test accuracy from 13.82% to 14.32% over uniform allocation across 10 matched seeds, and also improved worst-client accuracy. In matched-budget comparisons against representative partial-training baselines, HASA achieved the strongest worst-client and tail-client accuracy. An ablation study confirmed that the allocation direction matters: assigning smaller subnets to more heterogeneous clients — the reverse of HASA's approach — degraded both mean and tail performance. A cross-domain image-classification experiment further revealed that HASA's effectiveness depends on how accurately the heterogeneity score captures clients' actual need for greater model capacity.
What's missing
The study is limited to small-scale benchmarks (seven clients for the primary NLP task), and it is unclear how HASA scales to larger federated networks with hundreds or thousands of clients. Privacy implications of computing heterogeneity scores from local data — and whether this process itself leaks information — are not discussed.
What different sources said
- arXiv cs.AICenter
EvoCSFL: Surrogate-Assisted Evolutionary Client Selection for Efficient and Robust Federated Learning
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.