QSplitFL: New Deep Learning Framework Optimizes Model Training on Resource-Constrained Devices
Researchers have proposed QSplitFL, a Deep Q-Network-based framework that dynamically selects optimal split points in Split Federated Learning systems based on client hardware capabilities. The work addresses a persistent challenge in privacy-preserving distributed machine learning: fixed model split points can overload weaker devices, slowing convergence and reducing training stability across heterogeneous hardware. The framework, accepted at ECML-PKDD 2026, could improve the practicality of federated learning deployments on edge devices with varying computational resources.
QSplitFL introduces a capability-aware reinforcement learning approach to Split Federated Learning (SFL), where a Deep Q-Network (DQN) agent learns to select the layer at which a neural network model is divided between a client device and a central server. Rather than relying on high-dimensional model weight representations as state inputs, the framework uses lightweight hardware metrics—CPU utilization, memory, battery level, and network latency—making it computationally feasible on constrained devices. A decayed loss-drop reward function is designed to prioritize early convergence, and a committee-based DQN architecture with majority voting is introduced to reduce reward hacking, a common failure mode in reinforcement learning systems. Experiments were conducted across four datasets (MNIST, Fashion-MNIST, CIFAR-10, CIFAR-100) and four neural network architectures (CNN, ResNet50, MobileNetV4, ConvNeXt), with results reported as superior convergence and accuracy compared to existing methods. The source code has been made publicly available, and the paper has been accepted for presentation at ECML-PKDD 2026.
What's missing
The abstract does not quantify the magnitude of accuracy or convergence improvements. Key open questions include scalability to very large numbers of heterogeneous clients, sensitivity of the DQN to the choice of hardware metrics, and whether privacy guarantees are formally analyzed or assumed from the underlying SFL protocol. The real-world communication overhead of the committee-based DQN architecture relative to simpler approaches is also not addressed in the abstract.
What different sources said
- arXiv cs.LGCenter
QSplitFL: Capability Aware Deep Q-Learning for Optimal Split Point Selection in Split 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.