A5103267643- Advanced Graph Neural Networks
- Bioinformatics and Genomic Networks
- Complex Network Analysis Techniques
- Text and Document Classification Technologies
- Antibiotic Resistance in Bacteria
- Antibiotics Pharmacokinetics and Efficacy
- Recommender Systems and Techniques
- Natural product bioactivities and synthesis
- Machine Learning and ELM
- Pharmacological Effects of Medicinal Plants
- Land Use and Ecosystem Services
- Smart Agriculture and AI
- Evolutionary Game Theory and Cooperation
- Environmental Changes in China
- Evolution and Genetic Dynamics
- Generative Adversarial Networks and Image Synthesis
- Connective tissue disorders research
- Human Mobility and Location-Based Analysis
- Mathematical Analysis and Transform Methods
- Facial Rejuvenation and Surgery Techniques
- Autonomous Vehicle Technology and Safety
- Image Enhancement Techniques
- Neural Networks and Reservoir Computing
- Chaos control and synchronization
- Pediatric Hepatobiliary Diseases and Treatments
Yangzhou University
2021-2024
Shaanxi University of Chinese Medicine
2023-2024
Wuhan Textile University
2024
Taiyuan Institute of Technology
2023
Huazhong University of Science and Technology
2002-2022
Huashan Hospital
2020
Fudan University
2019-2020
State Key Laboratory of Genetic Engineering
2019
Shanghai Center for Brain Science and Brain-Inspired Technology
2019
Shanghai Institute for Science of Science
2019
The prediction of drug-disease associations holds great potential for precision medicine in the era big data and is important identification new indications existing drugs. between drugs diseases can be regarded as a complex heterogeneous network with multiple types nodes links. In this paper, we propose method, namely HED (Heterogeneous Embedding Drug-disease association), to predict based on network. Specifically, constructed from known associations, employs embedding characterize then...
Heterogeneous graph neural networks (HGNNs) have demonstrated promising capabilities in addressing various problems defined on heterogeneous graphs containing multiple types of nodes or edges. However, traditional HGNN models depend label information and capture the local structural original graph. In this paper, we propose a novel Graph Contrastive Learning method with Augmentation (AHGCL). Specifically, construct an augmentation by calculating feature similarity to latent information. For...
Abstract Background Platelet‐rich plasma (PRP) is effective in the treatment of androgenetic alopecia (AGA). Aims The purpose this study to assess effect PRP on proliferation human follicle dermal papilla cells (HFDPCs), observe growth hair follicles and shaft vitro, measure factors, evaluate efficacy safety injection. Patients/Methods HFDPCs was observed. length vitro measured. Then, concentration factors (EGF, FGF‐2, FGF‐7, IGF‐1, HGF, PDGF‐BB, VEGF‐A) evaluated. Half‐head injection...
Network embedding is a promising field and important for various network analysis tasks, such as link prediction, node classification, community detection others. Most research studies on prediction focus simple networks pay little attention to hypergraphs that provide natural way represent complex higher-order relationships. In this paper, we propose method with using (HNE). HNE adapts traditional method, Deepwalk, in hypergraphs. Firstly, the hypergraph model constructed based...
It is a challenging work to assess research performance of multiple institutes.Considering that it unfair average the credit institutes which in different order from paper, this we present allocation method (CAM) with weighted coefficient for institutes.The results APS dataset 18987 show top-ranked obtained by CAM correspond well-known universities or labs high reputation physics.Moreover, evaluate when citation links are added rewired randomly quantified Kendall's Tau and Jaccard index.The...
Modeling and simulation are essential methods to better understand a complex system in the real world. Many systems have strong demand for involving physical devices or equipment improve fidelity. But most of current cloud based designed with adequate resources suffering from high latency their centralized communication. Edge computing is getting much attention as new paradigm distributed heterogeneous requiring real-time However, edge still its initial stage there no consensus on framework...
In this paper, we investigate the reconstruction of networks based on priori structure information by Element Elimination Method (EEM). We firstly generate four types synthetic as small-world networks, random regular and Apollonian networks. Then, randomly delete a fraction links in original Finally, employ EEM, resource allocation (RA) structural perturbation method (SPM) to reconstruct with 90% information. The experimental results show that, comparing RA SPM, EEM has higher indices...
Fashion trend forecasting has consistently remained a focal point within the realm of fashion. Existing methods predominantly concentrate on external factors influencing fashion trends, often disregarding intricate interplay among distinct elements, namely "spatial dependencies" them. It is also significant challenge to excavate complicated temporal relationships in complex time series data. In this research, our primary focus modeling diverse elements and dependencies First, we propose...
Abstract Objective This study aimed to develop and apply a nomogram with good accuracy predict the risk of CRAB infections in neuro-critically ill patients. In addition, difficulties expectations application such tool clinical practice was investigated. Methods A mixed methods sequential explanatory design utilized. We first conducted retrospective identify factors for development patients; further validate predictive model. Then, based on developed tool, medical staff neuro-ICU were...
Traditional Graph Neural Network (GNN), as a graph representation learning method, is constrained by label information. However, Contrastive Learning (GCL) methods, which tackles the problem effectively, mainly focus on feature information of global or small subgraph structure (e.g., first-order neighborhood). In paper, we propose Local Structure-aware method (LS-GCL) to model structural nodes from multiple views. Specifically, construct semantic subgraphs that are not limited neighbors. For...
Patients in neuro-ICU are at a high risk of developing nosocomial CRKP infection owing to complex conditions, critical illness, and frequent invasive procedures. However, studies focused on constructing prediction models for assessing the neurocritically ill patients lacking present. Therefore, this study aims establish simple-to-use nomogram predicting admitted neuro-ICU. Three easily accessed variables were included model, including number antibiotics used, surgery, length stay. This might...