graph neural networks

Integrating Artificial Intelligence and Environmental Metagenomics for Ecosystem Monitoring and Management

Artificial Intelligence (AI) is driving a significant transformation in microbiology and environmental metagenomics, shifting the field from a descriptive framework towards a predictive and systems-level understanding. The advent of metagenomics has enabled direct analysis of microbial communities from environmental sa...

Individual brain similarity networks across aging and the Alzheimer's disease continuum

Aging is accompanied by gradual changes in brain structure and cognition, but these changes do not unfold in the same way for every individual. Some people maintain cognitive function into late life, whereas others experience decline in memory, executive function, processing speed, or other cognitive domains. This vari...

Multi-depth Rubidium PET Graph Neural Network for Major Adverse Cardiac Event Prediction

RumorGAL: Graph-anchored LLM for domain-generalized rumor detection

Rumor detection models must remain robust to breaking events that are unseen during training. Although large language models (LLMs) offer strong generalization potential, their text-centric nature makes them vulnerable to spurious textual patterns in noisy social media discussions and leaves them without explicit aware...

DrugVision: An AI-Based Clinical Solution for Medication Safety and Drug Interaction Detection.

DrugVision is an AI-based clinical decision support platform designed to enhance medication safety and detect adverse drug interactions. The system integrates a convolutional neural network ensemble (ResNet-50, EfficientNet-B0, MobileNetV3) for solid oral dosage form identification, transformer-augmented OCR for handwr...

Domain-Adversarial Disentanglement and Physics Constraints Enable Robust Structural Damage Identification Under Environmental Variability

Structural damage identification under environmental and operational variations (EOVs) remains a major challenge in civil infrastructure health monitoring because environmental shifts can be statistically correlated with damage-sensitive vibration features. This study presents a physics-informed spatiotemporal graph ne...

Asymmetric Focal Loss Improves Graph Neural Network Prediction of Drug-Pair-Associated Polypharmacy Side Effects

Accurate prediction of drug-pair-associated polypharmacy side effects remains an important challenge in computational pharmacovigilance. Standard binary cross-entropy (BCE) may provide insufficient emphasis on difficult positive examples. We evaluated ClinicalFocal, an asymmetric focal-loss function that assigns differ...

Cross-Radical Knowledge Sharing via Graph Neural Networks for Unified Prediction of Aqueous Organic Contaminant Second-Order Radical Reaction Rate Constants

Abstract Selecting among hydroxyl (HO•), sulfate (SO4•–), and carbonate (CO3•–) radical advanced oxidation processes requires reliable intrinsic second-order rate constants, yet the corresponding curated data sets contain 1250, 493, and 248 records, respectively. Reframing this imbalance as a cross-radical few-shot lea...

Operating-environment risk identification for dangerous goods transport vehicles based on unsafe driving behaviors

Operating environments, including road infrastructure, traffic flow, weather conditions, and mileage, directly influence driving behavior. Because driving behavior ultimately determines road traffic safety, it is critical to determine whether operating environments create conditions that may induce unsafe driving pract...

PEGNet: A Peridynamics-Inspired and Emergent-Feature-Conditioned Spatio-Temporal Graph Neural Network for Land Subsidence Modeling

Land subsidence prediction remains challenging. Conventional grid-based or sequence-only neural networks struggle to represent these spatial dependencies and often lack structured mechanisms for incorporating region-level deformation priors and local physical consistency. This study develops PEGNet, a Peridynamics-insp...

Weights Learn Experiences, Structures Crystallize Knowledge: Trace-Governed Structural Plasticity for Continual Learning — A Controlled Proof-of-Mechanism Study

This record contains the manuscript “Weights Learn Experiences, Structures Crystallize Knowledge: Trace-Governed Structural Plasticity for Continual Learning — A Controlled Proof-of-Mechanism Study”. The paper investigates whether repeated explicit reasoning can alter a learner’s computation graph, not only its numeric...

Dynamic Skill Demand Prediction for New Energy Manufacturing Using a Knowledge Graph-Enhanced Spatiotemporal Graph Neural Network

Accurate and forward-looking skill demand prediction is critical for workforce planning, curriculum development, and talent cultivation in the rapidly evolving new energy manufacturing sector. Existing approaches often rely on static occupational categories, retrospective skill taxonomies, or purely temporal models and...