Browse Articles

A universal and accurate real-time PCR assay for psittacine sex determination

Diagnostic accuracy of CRP and myeloperoxidase as a 2-marker biosignature for the diagnosis of symptomatic TB

Physics-informed-GPR data augmentation framework for flower-shaped antenna performance optimization and prediction using machine learning models

Morphology-keyed secure representation learning for privacy-preserving ECG arrhythmia classification and signal recovery

AE responses, energy evolution, and damage constitutive model of jointed rock-like material subjected to unilaterally confined compression

Decentralized federated distillation for privacy-preserving cross-league basketball data collaboration

Abstract Cross-league basketball analytics promises richer, more transferable performance models, yet competitive sensitivities and data-protection regulations make raw data sharing across leagues impractical. We propose a decentralized federated distillation framework that lets multiple basketball leagues co-train predictive models without centralizing data and without depending on a trusted aggregator. Each league node trains a locally chosen model on its anonymized aggregate game-level statistics and exchanges only temperature-scaled soft predictions with neighbors over a sparse peer-to-peer graph. To address re-identification threats and cross-league feature-space mismatch, the pipeline pairs an ε-differential-privacy Laplace mechanism applied directly to the released soft predictions—with explicit Rényi-DP composition across rounds—with k-anonymity for quasi-identifier coarsening and a Wasserstein optimal-transport projection that aligns league-specific feature spaces into a shared 64-dimensional representation. We establish convergence guarantees for federated distillation over decentralized communication graphs under non-convex objectives and heterogeneous data distributions, deriving an explicit bound that exposes the joint role of network spectral gap, distillation approximation error, transport-alignment error, and data heterogeneity. On a four-league dataset spanning the NBA, CBA, EuroLeague, and KBL—33,048 games in total—the proposed method attains 78.4% game-outcome accuracy, only 1.8 points behind a centralized oracle, while cutting communication overhead by more than 98% relative to parameter-averaging alternatives and preserving formal differential-privacy guarantees. Ablation studies confirm that feature alignment and adaptive temperature scheduling are both indispensable, and the sparse custom topology balances convergence speed against bandwidth efficiency.

Past subarctic marine food web shifts recovered by sedaDNA and network analysis

Abstract Subarctic marine ecosystems are highly sensitive to current climate and sea-ice change but long-term evidence for food web shifts is lacking. Sedimentary ancient DNA (sedaDNA) data covering the last 124,000 years combined with network analysis provides evidence for a shift from glacial (GLC) bottom-up to a deglacial-interglacial (IG) top-down food web structure. A consensus network approach, calibrated on generalized Lotka-Volterra simulations, delivers distinct modules that potentially resemble interglacial and glacial trophic structures. The glacial food web with a dominance of sea-ice adapted primary producers, densely linked with each other, supports a high module robustness and resource-driven bottom-up control. In contrast, the deglacial-interglacial food web with predominant consumers, like salmon, herring and small whales, are trophically diverse but rather fragile. Our study assumes that mobile consumers are facilitated by habitat expansion through submerged shelves and ice-free conditions during deglacial-interglacial periods altering the composition and structure of higher trophic levels. Our study implies that future warming, sea-ice decline and sea level rise may affect the structure and stability of subarctic consumer-driven, top-down food webs with adverse consequences for ecosystem services.

Profiling fungal secondary metabolites with anti-leishmanial activity: an integrated in silico and in vitro study

Multimodal fault diagnosis method of bogie motor based on physics-inspired regularization and enhanced ConvNeXt

Mechanical and microstructural properties of concrete incorporating high-strength graphene oxide-based geopolymer aggregate

Investigation of nano-particle effects on cold storage performance using finite element modeling

Design and additive manufacturing of 3D auxetic octahedron structure for enhanced specific energy absorption

Data-augmented machine learning approach for determination of over-excavation criteria in earth pressure balance shield tunnel boring machine operations

Team-based versus individual learning for out-of-hospital cardiac arrest rhythm recognition in newly hired emergency medical service providers: cluster-randomized equivalence trial

An integrated dynamic-evaluation and multi-task deep learning framework for load forecasting and compressor scheduling in nonlinear industrial air systems

Sensory-driven neck–limb coordination mechanisms for walk-trot-gallop gait transitions

Dynamic structured pruning and merging for real-time acceleration of reinforcement learning

Dyadic coping profiles and emotional distress in Chinese infertile couples

Holistic oncologic management improves quality of life in driver gene negative advanced NSCLC a latent growth curve modeling study

Chemo selective synthesis and antibacterial assessment of pyrazolo[3,4-d]pyrimidinones via a novel urea/thiourea-based ternary DES