A5113411156- Advanced Graph Neural Networks
- Recommender Systems and Techniques
- Chronic Obstructive Pulmonary Disease (COPD) Research
- Atomic and Subatomic Physics Research
- Dementia and Cognitive Impairment Research
- Mental Health Treatment and Access
- Cardiac Health and Mental Health
- Lung Cancer Treatments and Mutations
- Cardiovascular Function and Risk Factors
- Lung Cancer Research Studies
- Brain Tumor Detection and Classification
- Blood Pressure and Hypertension Studies
- Pulmonary Hypertension Research and Treatments
- Cancer Immunotherapy and Biomarkers
- Neural Networks and Applications
Guangzhou University of Chinese Medicine
2025
Meizhou City People's Hospital
2024
University of Hong Kong
2024
Integrated Chinese Medicine (China)
2019
Heterogeneous Graph Neural Networks (HGNNs) leverage diverse semantic relationships in Graphs (HetGs) and have demonstrated remarkable learning performance various applications. However, current distributed GNN training systems often overlook unique characteristics of HetGs, such as varying feature dimensions the prevalence missing features among nodes, leading to suboptimal or even incompatibility with HGNN training. We introduce Heta, a framework designed address communication bottleneck...
Dilated cardiomyopathy (DCM) is characterized by unilateral or bilateral ventricular enlargement and reduced systolic function, with without heart failure. In previous studies, we found that a history of chronic obstructive pulmonary disease (COPD) bronchitis high risk factor for DCM combined hypertension (PH). Therefore, propose the comorbidity COPD will increase cardiogenic mortality patients DCM.
Graph Neural Networks (GNNs) play a crucial role in various fields. However, most existing deep graph learning frameworks assume pre-stored static graphs and do not support training on streams. In contrast, many real-world are dynamic contain time domain information. We introduce GNNFlow, distributed framework that enables efficient continuous temporal representation multi-GPU machines. GNNFlow introduces an adaptive time-indexed block-based data structure effectively balances memory usage...